WO2025127558A1 - Drone and drone control method therefor - Google Patents

Abstract

The present embodiments provide a drone and a drone control method therefor, the drone being controlled to fly along a specific flight trajectory through the conversion of a motor signal into a motor control signal under normal circumstances, using a trained spike neural network to measure orientation on the basis of sensor data so that specific situations are recognized, and being controlled to fly while maintaining orientation through the adjustment of the motor control signal in the specific situations.

Description

Drone and method of controlling the drone

The present embodiments relate to a drone for controlling an aircraft in a specific situation and a method for controlling the drone.

In drone control systems, attitude and speed are important factors that determine the stability and maneuverability of the drone.

Most drones pose a risk of drifting or tilting under certain conditions, such as when subjected to gusts of wind, which can lead to a crash or loss of control.

However, general drones do not have a way to control the drone so that it does not drift or tilt and crash or lose control in certain situations such as gusts, and when controlling the drone in certain situations such as gusts with a general control method, the high power consumption and slow computation speed make it unsuitable for application in gusts that require flight time, stability, and quick response.

The present embodiments can provide a drone and a method for controlling the drone that stably controls the drone with low power and low latency in specific situations such as a gust of wind.

The present embodiments provide a drone and a control method for the drone, which controls the drone to fly along a specific flight trajectory by converting a motor signal into a motor control signal in a general situation, recognizes a specific situation by measuring attitude based on sensor data using a learned spike neural network, and controls the drone to fly while maintaining attitude by adjusting the motor control signal in a specific situation.

In one aspect, the present embodiments may provide a drone including a drone sensor unit that provides sensor data, a drone flight unit that provides motor signals, and a drone control unit that converts the motor signals into the motor control signals in general situations to control the drone to fly along a specific flight trajectory, measures attitude based on the sensor data using a learned spike neural network to recognize a specific situation, and controls the drone to fly while maintaining attitude by adjusting the motor control signals in specific situations.

In another aspect, the present embodiments can provide a method for controlling a drone, including a first step of controlling the drone to fly along a specific flight trajectory by converting a motor signal into a motor control signal in a general situation, and a second step of controlling the drone to recognize a specific situation by measuring an attitude based on sensor data using a learned spike neural network, and to control the drone to fly while maintaining the attitude by adjusting the motor control signal in the specific situation.

According to the drone and the method for controlling the drone according to the present embodiments, the drone can be stably controlled with low power and low delay in specific situations such as a gust of wind.

Figure 1 is a front view of a drone according to one embodiment.

Figure 2 is a configuration diagram of the drone of Figure 1.

Figure 3 is a schematic diagram of a portion of a spike neural network included in the drone control unit of Figure 2.

Figure 4 is a configuration diagram of a drone control unit of a drone according to another embodiment.

Figure 5 is a configuration diagram of a drone control unit of a drone according to another embodiment.

Figure 6 is a conceptual diagram of the flight control transfer process of the conventional control microprocessor and the gust situation analog computer of Figure 5.

Figure 7 illustrates input/output signals of the analog computer of Figure 6.

Figure 8 is a flow chart of a drone control method according to another embodiment.

Figure 9 illustrates in detail the drone control processes of the second stage of Figure 8.

Hereinafter, some embodiments of the present disclosure will be described in detail with reference to exemplary drawings. When adding reference numerals to components of each drawing, the same components may have the same numerals as much as possible even if they are shown in different drawings. In addition, when describing the present embodiments, if it is determined that a specific description of a related known configuration or function may obscure the gist of the technical idea of the present invention, the detailed description thereof may be omitted. When “includes,” “has,” “consists of,” etc. are used in this specification, other parts may be added unless “only” is used. When a component is expressed in the singular, it may include a case where the plural is included unless there is a special explicit description.

Additionally, in describing components of the present disclosure, terms such as first, second, A, B, (a), (b), etc. may be used. These terms are only intended to distinguish the components from other components, and the nature, order, sequence, or number of the components are not limited by the terms.

In a description of the positional relationship of components, when it is described that two or more components are "connected", "coupled" or "connected", it should be understood that the two or more components may be directly "connected", "coupled" or "connected", but the two or more components and another component may be further "interposed" to be "connected", "coupled" or "connected". Here, the other component may be included in one or more of the two or more components that are "connected", "coupled" or "connected" to each other.

In the description of the temporal flow relationship related to components, operation methods, or manufacturing methods, for example, when the temporal chronological relationship or the chronological flow relationship is described as "after", "following", "next to", or "before", it can also include cases where it is not continuous, as long as "immediately" or "directly" is not used.

Meanwhile, when a numerical value or its corresponding information (e.g., level, etc.) for a component is mentioned, even if there is no separate explicit description, the numerical value or its corresponding information may be interpreted as including an error range that may occur due to various factors (e.g., process factors, internal or external impact, noise, etc.).

The embodiments are described in detail with reference to the drawings below.

Figure 1 is a front view of a drone according to one embodiment.

Referring to FIG. 1, a drone (100) according to one embodiment can fly using a propeller (110) and an electric motor (120). The electric motor (120) is used to convert electric power into electrical energy to rotate the propeller (110). This rotational motion pushes out air, creating a force that propels the drone (100) upward. The propeller (110) helps to move the air using this rotational motion, and the drone can fly using this principle.

Drones (100) are equipped with various sensors, cameras, GPS, communication systems, etc. and are used for flight control and data collection.

The drone (100) has multiple (e.g., four in FIG. 1) propellers (110), and each propeller (110) can be controlled to perform ascent, descent, forward, backward, left and right movement, rotation, etc. To this end, the direction and height of the drone (100) are controlled by controlling the rotation speed of each propeller (110).

The drone (100) has a built-in system for controlling flight. This system controls the speed of each electric motor (120) and adjusts the attitude to move in the desired direction according to the command entered by the user.

A drone (100) according to one embodiment stably controls attitude and speed without drifting, tilting, crashing or losing control with low power and low latency in specific situations such as gusts of wind.

Figure 2 is a configuration diagram of the drone of Figure 1.

Referring to FIG. 2, a drone (100) according to one embodiment includes a drone sensor unit (130) that provides sensor data, a drone flight unit (140) that provides a motor signal, and a drone control unit (150) that converts a motor signal into a motor control signal in a general situation and controls the drone to fly along a specific flight trajectory, measures attitude based on the sensor data using a learned spike neural network (SNN) to recognize a specific situation, and controls the drone to fly while maintaining attitude by adjusting the motor control signal in a specific situation.

Hereinafter, a specific situation is exemplified by a gust situation as described above, but is not limited thereto, and generally includes all cases other than a general situation in which a drone (100) may drift or tilt, crash, or lose control. That is, this specification defines two situations in which a drone (100) may perform a general control operation in a first situation such as a general situation, and perform a special control operation in a second situation such as a specific situation. As described below, the drone (100) may perform a software control operation or a hardware control operation in the general situation and the special situation.

The drone sensor unit (130) includes a gyro sensor (132) and an acceleration sensor (134), and the sensor data may be, but is not limited to, angular velocity sensed by the gyro sensor (132) and acceleration sensed by the acceleration sensor (134).

The gyro sensor (132) can measure angular velocity. An example of the gyro sensor (132) may be a gyroscope that measures angular velocity. The acceleration sensor (134) is a sensor that measures acceleration. The acceleration sensor (132) can measure acceleration in the x-axis, y-axis, and z-axis directions, for example.

The drone flight unit (140) can convert the tuned motor control signal using a learned spike neural network (SNN) into a specific PWM signal that adjusts the motor speed and direction.

The drone flight unit (140) converts the motor control signal generated from the spike neural network (SNN) into a corresponding PWM signal that adjusts the motor speed and direction. The control logic of the drone flight unit (140) compensates for wind disturbance and maintains the desired flight trajectory using aerodynamic and flight dynamics models.

The drone flight unit (140) provides the converted PWM signal to the electric motor (120) and controls the electric motor (120), thereby inducing the flight of the drone (200).

The drone control unit (150) includes a microprocessor (152), and the drone (100) may additionally include a memory (not shown). The memory stores various sensor data and various programs. The memory may be a volatile memory (e.g. SRAM, DRAM) or a nonvolatile memory (e.g. NAND Flash).

For example, the microprocessor (152) can control the flight along a specific flight trajectory by converting a motor signal into a motor control signal in a general situation. In addition, the microprocessor (152) can measure the attitude based on sensor data using a learned spike neural network (SNN) to recognize a specific situation, and control the flight while maintaining the attitude by adjusting the motor control signal in a specific situation.

As another example, a microprocessor (152) included in a drone control unit (150) may control the drone to fly along a specific flight trajectory by converting a motor signal into a motor control signal in a general situation, store a learned spike neural network (SNN) in a memory, and execute it in the microprocessor (152) to measure attitude based on sensor data using the spike neural network (SNN), recognize a specific situation, and control the drone to fly while maintaining attitude by adjusting a motor control signal in a specific situation.

As another example, the drone control unit (150) may include separate hardware that measures attitude based on sensor data using a learned spike neural network (SNN) to recognize a specific situation, and controls flight while maintaining attitude by adjusting motor control signals in a specific situation. The separate hardware may be implemented as an analog circuit or a semiconductor chip, and may be a digital circuit or an analog circuit. In this regard, it will be described later with reference to FIGS. 4 and 5.

A neural network is an artificial intelligence technique that imitates the neural network of the human brain and is applied in various industries such as data mining, language recognition, image processing, and signal processing. In the past, convolutional neural network techniques were widely used, but since image data must be extracted, compared, and all data must be stored, it requires a large amount of memory and high power consumption, making it difficult to apply to small systems. Therefore, a spiking neural network (SNN), which is similar to the human brain's learning and information processing method and uses a small amount of data, has emerged.

Figure 3 is a schematic diagram of a portion of a spike neural network included in the drone control unit of Figure 2.

Referring to FIG. 3, the spike neural network (SNN) included in the drone control unit (150) includes a plurality of layers (Layers, 210), and each layer (210) contains a plurality of neurons. The layer (210) represents a group of a plurality of neurons having the same directionality.

Based on an arbitrary layer, it includes a pre-synaptic neuron (220) and a post-synaptic neuron (250), and a synapse (230) connecting the pre-synaptic neuron (120) and the post-synaptic neuron (250).

Neurons existing in different layers (210) are connected and signals are transmitted by synapses (230). In a neural network, the neuron before a synapse (230) is a presynaptic neuron (220), and the neuron after the synapse (230) that is connected to the presynaptic neuron (220) through the synapse (230) is a post-synaptic neuron (250). A signal generated from a presynaptic neuron (220) and transmitted to a post-synaptic neuron (250) through the synapse (230) is called an input spike (240). An output spike (260) is a signal generated from a post-synaptic neuron (250).

SNN (Spiking Neural Network) is a neural network method in which an output spike (260) occurs only when the sum of input spikes (240) exceeds the firing threshold of the output neuron.

That is, the post-synaptic neuron (250) multiplies each spike input from the pre-synaptic neuron (220) by a synaptic weight, and when the sum of all these values exceeds a certain threshold, it fires to the next layer (210).

Therefore, we need to find the optimal weights so that they can exceed a certain threshold, and then adjust the amount of change in the weights to obtain new synaptic weights, and then repeat the above calculations to control the output.

In SNN, an unsupervised learning method called STDP (Spiking Timing Dependent Plasticity) is usually used. The STDP learning method can control the output by optimizing the synaptic weights.

In SNN, when an input spike (240) is transmitted to an output neuron (250) through a synapse (230), each synapse (230) has a synapse weight. The synapse weight is an arbitrary value that is multiplied by the input spike (240) to amplify or attenuate the input spike (240). The input spikes (240) multiplied by the synapse weight are added and compared with the neuron threshold of the output to control the output. Therefore, an optimized synapse weight design for controlling the output is required. STDP (Spiking Timing Dependent Plasticity) is a function that calculates the amount of change in the synapse weight by using the difference between the time when the input spike (240) occurs and the time when the output spike (260) occurs.

That is, the weight of the synapse is adjusted by the difference between the occurrence time of the input spike (240) in the post-synaptic neuron (250) and the occurrence time of the output spike (260) in the pre-synaptic neuron (220).

The drone control unit (150) applies an algorithm for learning and classifying a spike neural network (SNN). In the process of learning and classifying given data, the artificial neuron analyzes the attitude information of the drone (100) through the generation of a small number of spikes at multiple points simultaneously. Through this, the spike neural network (SNN) analyzes the attitude information of the drone (100) and generates a motor control signal.

A control signal can be generated based on the attitude information of the drone (100) analyzed by the spike neural network (SNN). Through this, the spike neural network (SNN) can quickly control the attitude of the drone (100).

As mentioned above, general drones do not have a way to control the drone in certain situations such as gusts to prevent it from drifting or tilting and crashing or losing control, and when controlling the drone in certain situations such as gusts with a general control method, the high power consumption and slow computation speed make it unsuitable for application in gusts that require flight time, stability, and quick response.

A drone (100) according to one embodiment solves this problem by using a spiking neural network (SNN) to adjust motor control signals based on sensor data, compensate for gust effects, and maintain a desired flight trajectory.

Spike neural networks (SNNs) have the advantage of low power and low latency compared to traditional networks such as deep learning and recurrent neural networks in terms of spatiotemporal information processing. These advantages include low power consumption, high parallel processing, and real-time operation. This is very important in drone control systems where low latency and low power consumption are essential for long-term and safe operation.

A drone (100) according to one embodiment provides a spike neural network-based drone optimized for low power and low latency operation in specific situations, such as gust conditions, for example. The drone (100) according to one embodiment can provide efficient and powerful control while minimizing power consumption by integrating sensor data, spike neural network, and motor control system as subsystems. In addition, the drone (100) according to one embodiment can optimize performance in various weather conditions through the adaptability and learning mechanism of the spike neural network.

Hereinafter, examples of implementing separate hardware to measure attitude based on sensor data using a learned spike neural network (SNN) to recognize a specific situation, and control flight while maintaining attitude by adjusting motor control signals in a specific situation are described with reference to FIGS. 4 and 5.

Figure 4 is a configuration diagram of a drone control unit of a drone according to another embodiment.

Referring to FIG. 4, in a drone (100A) according to another embodiment, a drone control unit (150) may include a first microprocessor (154) that converts a motor signal into a motor control signal in a general situation and controls the drone to fly along a specific flight trajectory, and a second microprocessor (156) that measures attitude based on sensor data using a learned spike neural network (SNN) to recognize a specific situation, and controls the drone to maintain the flight trajectory by adjusting the motor control signal in the specific situation.

The first microprocessor (154) may be a general microprocessor, and the second microprocessor (156) may be a dedicated microprocessor implementing a spike neural network (SNN). The second microprocessor (156) may be a gust situation control analog computer (158) that processes analog signals as described with reference to FIG. 5, but may also be a digital microprocessor that processes digital signals in the same manner as the first microprocessor (154).

Figure 5 is a configuration diagram of a drone control unit of a drone according to another embodiment.

Referring to FIG. 6, in a drone (100B) according to another embodiment, a drone control unit (150) may include a general control microprocessor (157) that converts a motor signal into a motor control signal in a general situation and controls the drone to fly along a specific flight trajectory, and a gust situation control analog computer (158) that measures attitude based on sensor data using a learned spike neural network (SNN) to recognize a specific situation, and controls the drone to fly while maintaining attitude by adjusting the motor control signal in a specific situation.

Typically, the control microprocessor (157) may be implemented as a von Neumann computer-based microprocessor system capable of performing flight and desired tasks of the drone (100) in general situations or normal situations rather than specific situations such as gusts of wind.

The drone control unit (150) recognizes a specific situation, such as a gust of wind, and transfers the drone's flight control to the SNN-based gust of wind situation control analog computer (158) based on the recognition.

As described above, the spike neural network (SNN) utilized in the gust situation control analog computer (158) is an artificial neural network that imitates the operating principles of biological neurons, and the spike neural network (SNN) is used to detect and analyze the attitude of the drone (100) and generate a control signal. The gust situation control analog computer (158) can implement a drone attitude control system with low power and low latency by applying the spike neural network (SNN) capable of processing analog signals.

The analog computer (158) for controlling the gust situation applies an algorithm for learning and classifying a spike neural network (SNN). In the process of learning and classifying given data, the artificial neuron analyzes the attitude information of the drone (100) through the occurrence of a small number of spikes at multiple points simultaneously. Through this, the spike neural network (SNN) analyzes the attitude information of the drone (100) and generates a motor control signal.

A device is required to generate a control signal based on the attitude information of the drone (100) analyzed by the spike neural network (SNN). Through this, the spike neural network (SNN) can quickly control the attitude of the drone (100).

A drone (100B) according to another embodiment described above utilizes a spike neural network (SNN) to quickly control and maintain the attitude of the aircraft in a gust of wind, and is designed using an analog computer-based integrated circuit for this purpose.

To this end, the attitude of the drone is quickly detected and analyzed through a spike neural network (SNN) to immediately generate a control signal, and a spike neural network (SNN) structure with low power and low latency is used for this purpose. Accordingly, the drone (100B) according to another embodiment described above can quickly and stably control the attitude of the drone (100B) in a gust of wind, thereby enabling safe flight.

Figure 6 is a conceptual diagram of the flight control transfer process of the conventional control microprocessor and the gust situation analog computer of Figure 5.

Referring to FIGS. 5 and 6, in a drone (100B) according to another embodiment, when a gust of wind situation is recognized, the drone control unit (150) may transfer control of the aircraft to a gust of wind situation control analog computer (158), and when the gust of wind situation is over, the control of the aircraft may be transferred to a normal control microprocessor (157).

Specifically, when the control of the aircraft is in the gust situation control analog computer (158), the determination of the degree of stabilization of the aircraft of the drone (100) is made through a spike neural network (SNN), and when the spike neural network (SNN) recognizes the end of the gust situation, the control is transferred to the existing conventional control microprocessor (157).

A drone (100B) according to another embodiment can more efficiently implement a spike neural network (SNN) using an analog computer-based integrated circuit. This is also called a neural processing unit or neuromorphic chip, and can more efficiently implement a drone attitude control system.

Figure 7 illustrates the input/output signals of the gust situation control analog computer of Figure 6.

Referring to FIG. 7, the gust situation control analog computer (158) can receive a motor signal and sensor data, and output a gust situation judgment signal and a motor control signal that judge a gust situation using a learned spike neural network (SNN).

The gust situation control analog computer (158) receives motor signals and sensor data, measures attitude based on the sensor data using a learned spike neural network (SNN), recognizes a specific situation, and outputs a gust situation judgment signal and a motor control signal in a specific situation to control flight while maintaining attitude.

The drone (100, 100A, 100B) according to the above-described embodiments can be recognized as superior to the conventional drone attitude control system. This is because it can quickly detect and analyze the surrounding environment and conditions, such as the attitude of the drone, the degree of influence due to wind, and prediction of future wind strength, using a spike neural network (SNN), and immediately generate a control signal while having low power and low latency, thereby showing more efficient performance than the conventional system.

The drone (100, 100A, 100B) according to the above-described embodiments has the advantage of being able to quickly and stably control attitude in a gust of wind situation, thereby enabling safe flight. Therefore, the drone (100, 100A, 100B) according to the above-described embodiments provides a high level of safety, stability, and performance in the drone industry.

Figure 8 is a flow chart of a drone control method according to another embodiment.

Referring to FIG. 8, a drone control method (300) according to another embodiment includes a first step (S310) of controlling the drone to fly along a specific flight trajectory by converting a motor signal into a motor control signal in a general situation, and a second step (S320) of measuring attitude based on sensor data using a learned spike neural network to recognize a specific situation, and controlling the drone to fly while maintaining attitude by adjusting the motor control signal in the specific situation.

In the second step (S320), the learned spike neural network can be used to convert the adjusted motor control signal into a specific PWM signal that adjusts the motor speed and direction.

As described above, the sensor data may be, but is not limited to, angular velocity sensed by a gyro sensor and acceleration sensed by an acceleration sensor.

As described above with reference to FIG. 3, the spike neural network (SNN) includes a plurality of layers (210), and each layer (210) includes a plurality of neurons, and based on an arbitrary layer (210), includes a pre-synaptic neuron (220) and a post-synaptic neuron (250), and a synapse (230) connecting the pre-synaptic neuron (220) and the post-synaptic neuron (250), and the weight of the synapse can be adjusted based on the difference in the occurrence time of an output spike (260) in the post-synaptic neuron (250) and the occurrence time of an input spike (240) in the pre-synaptic neuron (220).

For example, in the first step (S310), the first microprocessor (154) illustrated in FIG. 4 converts the motor signal into a motor control signal in a general situation and controls the flight to a specific flight trajectory, and in the second step (320), the second microprocessor (156) illustrated in FIG. 4 measures the attitude based on sensor data using a learned spike neural network (SNN) to recognize a specific situation, and controls the flight to maintain the attitude by adjusting the motor control signal in the specific situation.

As another example, in the first step (S310), the conventional control microprocessor (157) illustrated in FIG. 5 converts a motor signal into a motor control signal in a general situation and controls the flight to a specific flight trajectory, and in the second step (S320), the gust situation control analog computer (158) illustrated in FIG. 5 measures the attitude based on sensor data using a learned spike neural network to recognize a specific situation, and controls the flight to maintain the attitude by adjusting the motor control signal in the specific situation.

In the second step (S320), as illustrated in FIG. 6, if a gust situation is recognized, control of the aircraft may be transferred to a gust situation control analog computer (158), and if the gust situation has ended, control of the aircraft may be transferred to a normal control microprocessor (157).

As described above with reference to FIG. 7, the gust situation control analog computer (158) can receive motor signals and sensor data, and output a gust situation judgment signal and a motor control signal that judge a gust situation using a learned spike neural network.

Figure 9 illustrates in detail the drone control processes of the second stage of Figure 8.

Referring to FIG. 9, in the second step (S320) described above with reference to FIG. 8, the drone control right is transferred in a gust situation based on a motor signal (PWM) (S321), the input motor signal (PWM) and sensing data on the current state of the aircraft are input (S322), spike embedding and a spike neural network for the sensing data are input (S323), a PWM control signal suitable for each electric motor (220) is generated based on the input information and input to the electric motor (220) (S324), and after controlling the drone through the PWM control signal, the control right is terminated when the gust situation is resolved (S325).

As described above with reference to FIG. 6, when a gust situation is recognized, control of the aircraft can be transferred to a gust situation control analog computer (158), and when the gust situation has ended, control of the aircraft can be transferred to a normal control microprocessor (157).

The drone control method (300) according to another embodiment described above can be recognized as superior to conventional drone attitude control systems. It can quickly detect and analyze the surrounding environment and conditions, such as the attitude of the drone, the degree of influence by the wind, and prediction of future wind strength, using a spike neural network (SNN), and generate a control signal immediately while having low power and low latency, so that it can exhibit more efficient performance than conventional systems.

The drone control method (300) according to another embodiment described above has the advantage of being able to quickly and stably control attitude in a gust of wind situation, thereby enabling safe flight. Therefore, the drone control method (300) according to another embodiment described above provides a high level of safety, stability, and performance in the drone industry.

The above description is merely an example of the technical idea of the present disclosure, and those skilled in the art to which the present disclosure pertains may make various modifications and variations without departing from the essential characteristics of the technical idea of the present disclosure. In addition, the present embodiments are not intended to limit the technical idea of the present disclosure but to explain it, and therefore the scope of the technical idea of the present disclosure is not limited by these embodiments. The protection scope of the present disclosure should be interpreted by the claims below, and all technical ideas within a scope equivalent thereto should be interpreted as being included in the scope of the rights of the present disclosure.

CROSS-REFERENCE TO RELATED APPLICATION

This patent application claims the benefit of priority under 35 U.S.C. § 119(a) to Korean patent application No. 10-2023-0179317, filed December 12, 2023, the entire contents of which are incorporated herein by reference. In addition, this patent application claims priority for other countries than the United States for the same reasons, the entire contents of which are incorporated herein by reference.

Claims (16)

Drone sensor section that provides sensor data; A drone flight section providing motor signals; and A drone including a drone control unit that controls the drone to fly along a specific flight trajectory by converting the motor signal into the motor control signal in a general situation, measures the attitude based on the sensor data using a learned spike neural network to recognize a specific situation, and controls the drone to fly while maintaining the attitude by adjusting the motor control signal in the specific situation. In the first paragraph, The drone flight section is a drone that converts the motor control signal adjusted using the learned spike neural network into a specific PWM signal that adjusts the motor speed and direction. In the first paragraph, The above drone sensor section includes a gyro sensor and an acceleration sensor, and the sensor data is a drone, which is an angular velocity sensed by the gyro sensor and an acceleration sensed by the acceleration sensor. In the first paragraph, The above spike neural network includes a plurality of layers, and each layer includes a plurality of neurons, and based on an arbitrary layer, includes a pre-synaptic neuron, a post-synaptic neuron, and a synapse connecting the pre-synaptic neuron and the post-synaptic neuron, and a drone that learns to adjust the weight of the synapse based on the difference between the spike occurrence time in the post-synaptic neuron and the spike occurrence time in the pre-synaptic neuron. In the first paragraph, The drone control unit comprises a first microprocessor that converts the motor signal into the motor control signal in the general situation and controls the drone to fly in a specific flight trajectory, and a second microprocessor that measures the attitude based on the sensor data using the learned spike neural network to recognize a specific situation, and controls the drone to fly while maintaining the attitude by adjusting the motor control signal in the specific situation. In the first paragraph, The drone control unit includes a general control microprocessor that converts the motor signal into the motor control signal in the general situation and controls the drone to fly in a specific flight trajectory, and a gust situation control analog computer that measures the attitude based on the sensor data using the learned spike neural network to recognize a specific situation, and controls the drone to fly while maintaining the attitude by adjusting the motor control signal in the specific situation. In Article 6, A drone in which the above drone control unit, when recognizing the above gust situation, transfers control of the aircraft to the gust situation control analog computer, and when the gust situation ends, transfers control of the aircraft to a normal control microprocessor. In Article 6, The above wind gust situation control analog computer receives the motor signal and the sensor data, and outputs a wind gust situation judgment signal and the motor control signal using a learned spike neural network to judge the wind gust situation, the drone. A first stage for controlling the flight along a specific flight trajectory by converting a motor signal into a motor control signal in a general situation; and A method for controlling a drone, comprising a second step of measuring a posture based on the sensor data using a learned spike neural network to recognize a specific situation, and controlling the drone to fly while maintaining the posture by adjusting the motor control signal in the specific situation. In Article 9, A method for controlling a drone in the second step, wherein the motor control signal adjusted using the learned spike neural network is converted into a specific PWM signal for adjusting the motor speed and direction. In Article 9, The above sensor data is a method for controlling a drone, which is an angular velocity sensed by a gyro sensor and an acceleration sensed by an acceleration sensor. In Article 9, The above spike neural network includes a plurality of layers, and each layer includes a plurality of neurons, and based on an arbitrary layer, includes a pre-synaptic neuron, a post-synaptic neuron, and a synapse connecting the pre-synaptic neuron and the post-synaptic neuron, and a drone control method that learns to adjust the weight of the synapse based on the difference between the spike occurrence time in the post-synaptic neuron and the spike occurrence time in the pre-synaptic neuron. In Article 9, In the above first step, the first microprocessor converts the motor signal into the motor control signal in the above general situation and controls the flight to a specific flight trajectory, A method for controlling a drone, wherein in the second step, the second microprocessor measures the attitude based on the sensor data using the learned spike neural network to recognize a specific situation, and controls the drone to fly while maintaining the attitude by adjusting the motor control signal in the specific situation. In Article 9, In the above first step, the normal control microprocessor converts the motor signal into the motor control signal in the above general situation and controls the flight to a specific flight trajectory. A method for controlling a drone, wherein in the second step, a gust situation control analog computer measures the attitude based on the sensor data using the learned spike neural network to recognize a specific situation, and controls the drone to fly while maintaining the attitude by adjusting the motor control signal in the specific situation. In Article 14, A method for controlling a drone, wherein in the second step, if the gust situation is recognized, control of the aircraft is transferred to the gust situation control analog computer, and if the gust situation ends, control of the aircraft is transferred to a normal control microprocessor. In Article 14, A method for controlling a drone, wherein the above-mentioned wind gust situation control analog computer receives the above-mentioned motor signal and the above-mentioned sensor data, and outputs a wind gust situation judgment signal and the above-mentioned motor control signal using a learned spike neural network.

Images