On-line simulation system for unmanned mining operation
Technical Field
The invention belongs to the technical field of unmanned and system simulation, and particularly relates to an online simulation system for unmanned mining operation.
Background
The automatic driving technology is developed at a high speed, and unmanned driving under specific scenes is realized at home and abroad. The mining area is one of the scenes most suitable for falling of the unmanned technology due to the fact that the transportation road is closed, the transportation route is relatively fixed, and the speed of the operation vehicle is relatively low; the state also makes some relevant plans for the wisdom mine, wherein unmanned can solve the problem that the mining area transportation efficiency is low, the traffic accident is high. On the other hand, as an emerging industry, a mine unmanned simulation verification system for mine mining operation is not available at present, and a strict and lengthy development and verification process still needs to be performed when the mine unmanned simulation verification system reaches a fully mature technology and falls to the ground delivery stage.
The unmanned mining operation test in the mining area mainly comprises 3 types: mine sports car test, test field sports car test and simulation test. Because the speed of the roadster in the mining area is low, the road gradient is large, the number of curves is large, the roadster test efficiency is low, the accident is easy to happen, the real road condition and the operation scene of the mining area cannot be reproduced in the test of a test field, so that the online simulation verification system of the mining area is a necessary method for verifying the functions of the unmanned and dispatching platform and evaluating the mining operation efficiency, and the problems of long time consumption, high cost, difficult scene construction and the like of the actual roadster can be solved.
In mining operation, real-time scheduling control of mining equipment is essential for ensuring the safety and high efficiency of vehicle operation. In the unmanned system of the mining area, the dispatching platform plays a role in monitoring and dispatching vehicles. The mine card and the platform server realize real-time scheduling management through Transmission Control Protocol (TCP) communication. Although the existing system simulation software can realize the simulation of the operation process, simulation data cannot be uploaded in real time, so that the verification of the effectiveness of the unmanned communication function and the platform scheduling function cannot be realized, and the important technical problem of the operation simulation in a mining area is how to realize the communication between the unmanned simulation system and an external platform.
The mine unmanned mining operation system is a multi-agent system (MAS) and the equipment mainly comprises mine cards and auxiliary vehicles (electric shovels and unloading points). The mining area transportation road is different from a common urban road, does not have road marks such as traffic signs, indicator lights and lane lines, and has a road right concept for ensuring driving safety, namely whether vehicles are allowed to pass through the road section currently. The position of the transportation road changes frequently along with the mining of materials, the on-line simulation needs to model the multi-agent and adapt to the updating of map data in real time, and the difficulty is brought to the building of a simulation model. The existing unmanned driving and system simulation software, such as analog, Vissim and the like, aims at simulating urban and highway roads, cannot completely meet the simulation requirements of mining scenes, and needs an online simulation verification system to solve the defects.
Disclosure of Invention
In order to solve the existing technical problems, the invention provides an online simulation system for unmanned mining operation, which has the following specific scheme:
the system comprises an application layer, a service logic layer, a data access layer and a network transmission layer, wherein the four layers are mutually linked and are respectively kept relatively independent; the online simulation system is established on a platform, and the platform also comprises a platform server;
the application layer comprises a GUI module, a function verification module and an efficiency evaluation module, the GUI module is responsible for simulation interface setting and used for simulation interface display and man-machine interaction and can reflect simulation animations obtained by operation of the service logic layer in real time, the function verification module defines a function verification method of the unmanned vehicle end and the platform, simulation data output by the service logic layer are used and compared with internal logic to judge results, for complex scenes, the simulation initialization state is constructed to verify the effectiveness of functions in all scenes, logs and results of the function verification module are stored in the data access layer and can be simulated and reproduced by calling data stored in the data access layer, and the efficiency evaluation module defines an operation efficiency evaluation system of unmanned vehicles in a mining area and is used for statistical analysis of the simulation data; the business logic layer is realized based on an analog simulation platform and is used for controlling the operation flow of each device in a mining area, basic information required by simulation operation is acquired through the data access layer and the CUI module, a simulation scene is automatically modeled, device parameter data is converted into parameters of a simulation model, and communication with the platform server is realized by connecting the network transmission layer; the data access layer is responsible for establishing connection between each database and the service logic layer, inputting basic data necessary for simulation model operation and storing and outputting simulation results; and the network transmission layer is responsible for real-time communication between the service logic layer and the platform server.
The efficiency evaluation module comprises operation efficiency statistics and scheduling algorithm optimization, wherein the operation efficiency statistics specifically comprise statistics of quantitative evaluation indexes such as passing rate of each road section in a mining area, average queuing duration of each operation area, single goods pulling operation time of a mine truck, unit time utilization rate of auxiliary vehicles and the like, and an analysis icon is output to visually reflect mining operation efficiency; the scheduling algorithm optimizes an AI scheduling algorithm of an integrated platform and adopts an RL4J deep reinforcement learning package of Skymin to perform simulation experiments, so that the optimal scheduling strategy suitable for the mining area can be determined, and the efficiency of platform scheduling strategy development is greatly improved.
The multi-agent modeling and discrete event modeling method provided by the analog simulation platform can be used for multi-agent system modeling in a mining area, and the provided process modeling library can be used as a basis for drawing a work flow diagram and is used for controlling the work flow of each device in the mining area.
The business logic layer is also provided with an intelligent agent framework which specifically comprises a main intelligent agent, a mine card intelligent agent and an auxiliary vehicle intelligent agent; the main intelligent agent is responsible for initializing the simulation model established by the analog simulation platform, controlling the connection interaction of each intelligent agent and realizing the complex simulation of the multi-intelligent-agent system; the main intelligent agent obtains basic parameter information required by simulation operation through the data access layer and the GUI module, and automatically models a simulation scene by converting map data into geographic position information elements required by the simulation operation; the main agent is also responsible for simulation exception handling, initializes other agents according to requirements and transmits simulation parameters, and realizes the state control function of other agents through an application layer; the mine card intelligent agent starts the mining operation flow of the mine card intelligent agent after being initialized by the main intelligent agent, and a scene model constructed by the main intelligent agent runs; the intelligent mine card body realizes dynamic path planning, road right application and release and vehicle end logic of driving behavior control in an operation area through a decision planning module program integrated with an unmanned vehicle; the mine card intelligent body realizes real-time communication with the platform through a network transmission layer, and realizes communication with the auxiliary vehicle intelligent body through a scheduling platform information forwarding function; the auxiliary vehicle intelligent body comprises an electric shovel and a unloading point, is controlled by a mining operation flow, automatically interacts with an operation mine card and realizes microcosmic loading and unloading simulation;
the data access layer comprises a map database, an equipment parameter database and a simulation experiment result database.
The equipment parameter database mainly comprises an equipment name, an IMEI number, a mine card cargo capacity and an electric shovel bucket capacity; the map database is used for storing the geographical position information of the mining area by adopting an SQLite database, and is convenient to read and update in real time; the map database comprises information of each road section including geographic coordinates, information of control points and information of operation areas; the simulation output database is managed by an Access database and comprises simulation experiment process data, function verification result data and efficiency evaluation index data.
The network transmission layer comprises an HTTP module, a WebSocket module and a TCP client management module; the HTTP module is responsible for calling a platform rear-end interface and downloading updated map data, the WebSocket module is responsible for monitoring platform alarm events, and the TCP module is responsible for generating and analyzing messages according to the unmanned vehicle-end communication protocol.
The network transmission layer restores the actual communication scene to the maximum extent and can verify the gateway function by integrating the V2X communication module program at the unmanned vehicle end; the functions mainly comprise downloading the latest map data, downloading a scheduling task file, automatically processing platform scheduling alarm information, and communicating with the platform in real time to receive scheduling information; the TCP client management module initializes a gateway program for each agent and performs centralized management, so as to ensure that a single agent establishes stable and effective TCP long connection with the platform.
Compared with the prior art, the invention has the following beneficial effects:
1. the invention can simulate the unmanned mining operation flow under the scene of the mining area, verify the unmanned function, objectively describe the complex scheduling and road right control strategy of the mining area, and provide decision support for the setting of the attributes of the mining area road and the optimization of the scheduling and road right control strategy, wherein the user can visually see the efficiency of the mining operation system through a simulation interface.
2. The invention does not depend on scene modeling in advance, can automatically generate the mine scene according to the map database, can construct virtual map data by a user to overcome the limitation of actual mine sites, has strong applicability, can be directly operated when being deployed in any mine, and reduces the scene modeling and maintenance cost.
3. The invention is an online simulation system, the complete operation data generated by simulation can be stored in a dispatching platform, a large amount of data support is provided for development and verification of platform functions, and reference can be provided for online simulation of the unmanned system in other closed scenes.
Drawings
FIG. 1 is a block diagram of an unmanned mining operation on-line simulation verification system of the present invention
Detailed Description
The following detailed description of the present invention is provided in conjunction with the accompanying drawings, but it should be understood that the scope of the present invention is not limited to the specific embodiments.
As shown in FIG. 1, the unmanned mining operation online simulation verification system comprises an application layer, a business logic layer, a data access layer and a network transmission layer, wherein the four layers are mutually linked and are respectively kept relatively independent. The application layer is responsible for setting a simulation interface, defining a function verification method of the unmanned vehicle end and the platform and defining an operation efficiency evaluation system of the unmanned vehicle in the mining area; the business logic layer realizes the online simulation of the complex mining operation flow of the mining area through the interaction among the intelligent agents; the data access layer is used as a background database of the service logic layer to store data, and stores basic information and simulation output data required by simulation operation in the service logic layer; the network transmission layer establishes connection between the service logic layer and the platform server to realize real-time communication.
The application layer is responsible for simulation function definition and realization of a man-machine interaction function and comprises a graphical interface (GUI) module, a function verification module and an efficiency evaluation module.
And the GUI module of the application layer is responsible for simulation interface display and man-machine interaction and can reflect simulation animations obtained by operation of the service logic layer in real time. Before the simulation experiment, the types of the mining areas can be selected to switch different mining areas for simulation; the number of the operation mine cards in the system can be changed, the operation state of the equipment can be changed, the material can be changed, the operation speed of the simulation system can be controlled and the like during simulation operation. The user's settings are transmitted to the service logic layer in real time, ensuring dynamic visualization of the simulation results.
The function verification module compares simulation data output by the service logic layer with internal logic to judge results, and verifies the effectiveness of functions in all scenes by constructing a simulation initialization state for complex scenes. And the functional stability of the unmanned system is verified for many times rapidly by controlling the running speed of the simulation system. The log and the result of the function verification module are stored in the data access layer and can be simulated and reproduced through the database.
The efficiency evaluation module is used for statistical analysis of simulation data, including job efficiency statistics and scheduling algorithm optimization. The operation efficiency statistics mainly comprise statistics of quantitative evaluation indexes such as passing rate of each road section of a mining area, average queuing time of each operation area, single cargo pulling operation time of a mine truck, unit time utilization rate of auxiliary vehicles and the like, and an analysis icon is output to visually reflect the mining operation efficiency. The scheduling algorithm is optimized through an AI scheduling algorithm of an integrated platform and simulation experiments are carried out by adopting an RL4J deep reinforcement learning package of Skymin, so that the optimal scheduling strategy suitable for the mining area can be determined, and the efficiency of platform scheduling strategy development is greatly improved.
The business logic layer is realized based on a multi-agent system (called analog for short) simulation platform, the provided multi-agent modeling and discrete event modeling method can be used for multi-agent system modeling in a mining area, and the provided process modeling library can be used as a basis for drawing a process flow diagram and is used for controlling the operation process of each device in the mining area. The agent framework includes a main agent, a mine card agent, and an auxiliary vehicle agent.
The main agent is responsible for initializing the simulation model, controlling the connection interaction of each agent and realizing the complex simulation of the multi-agent system. The main intelligent agent obtains basic parameter information required by simulation operation through the data access layer and the GUI module, and automatically models a simulation scene by converting map data into geographic position information elements required by the simulation operation. The main intelligent agent is also responsible for simulation exception handling, initializes other intelligent agents according to requirements and transmits simulation parameters, and the state control and other functions of other intelligent agents are realized through an application layer.
The mine card intelligent agent starts the mining operation flow of the mine card intelligent agent after the main intelligent agent is initialized, and a scene model constructed by the main intelligent agent runs. The intelligent mine card body integrates a decision planning module program of the unmanned vehicle to realize vehicle end logics such as dynamic path planning, road right application and release, driving behavior control and the like in the operation area. The intelligent mine card body realizes real-time communication with the platform through a network transmission layer and realizes communication with the auxiliary vehicle in the simulation system through a scheduling platform information forwarding function.
The auxiliary vehicle intelligent body comprises an electric shovel and a unloading point, is controlled by a mining operation flow, automatically interacts with an operation mine card and realizes microcosmic loading and unloading simulation.
The data access layer is responsible for establishing connection between the database and the service logic layer, and the managed data comprise equipment parameter data, map data and simulation output data.
The data access layer inputs basic data necessary for the operation of the simulation model for the service logic layer and stores and outputs the simulation result. The equipment parameter data mainly comprises an equipment name, an IMEI number, a mine card cargo capacity, an electric shovel bucket capacity and the like; the SQLite database is adopted to store the geographical location information of the mining area, and real-time reading and updating are convenient. The map data storage section includes information on each road section including geographical coordinates, information on control points, and information on work areas. And managing simulation output data by adopting the Access database, wherein the simulation output data mainly comprises simulation experiment process data, function verification result data and efficiency evaluation index data.
Each module of the network transmission layer is compiled by Java language and is responsible for the communication between the service logic layer and the platform server, and the supported communication protocols comprise TCP (transmission control protocol), HTTP (hyper text transport protocol) and WebSocket protocol.
The network transmission layer restores the actual communication scene to the maximum extent by integrating the V2X communication module program at the unmanned vehicle end, and can verify the gateway function. The functions of the method mainly comprise the steps of downloading the latest map data, downloading scheduling task files, automatically processing platform scheduling alarm information, communicating with the platform in real time to receive scheduling information and the like. The TCP client management module initializes a gateway program for each agent and performs centralized management, so as to ensure that a single agent establishes stable and effective TCP long connection with the platform. When the simulation experiment runs, the information which needs to be sent to the dispatching platform is processed and sent out through the gateway program according to the operation flow in the service logic layer, and after the gateway program receives the platform information, the analyzed result is transmitted to the service logic layer to continue the operation flow; when the map data are updated, the simulation system actively sends an HTTP request to the platform to acquire the latest map data and transmits the latest map data to the service logic layer for scene construction processing; when the operation state of the equipment changes, the network transmission layer receives scheduling alarm information pushed by the platform through the WebSocket client, and calls a platform rear-end interface to process according to the service logic.
When the unmanned mining operation online simulation verification system is implemented, a decision planning module, a gateway module and a platform scheduling function (the prior art of the existing unmanned vehicle end) of an unmanned vehicle end are used as objects to be tested, and the authenticity and the reliability of the verification system are greatly improved by integrating the actual unmanned vehicle end module program. The system fully combines the characteristics of unmanned mining operation in the mining area, realizes real-time updating of geographic position information of the mining area, has the advantage of strong practical applicability, realizes real-time communication between multiple intelligent agents and an intelligent scheduling platform, realizes simulation verification of operation processes in functional areas such as a scheduling function, a loading area, an unloading area and a parking area, realizes dynamic display of simulation parameter setting and statistical charts, generates a large amount of effective full-flow roadster data, and improves the development efficiency of the unmanned operation system in the mining area.
The above-mentioned embodiments are merely specific embodiments of the present invention, which are used to illustrate the technical solutions of the present invention, but not to limit the technical solutions, and the scope of the present invention is not limited thereto, although the present invention is described in detail with reference to the above-mentioned embodiments. It should be noted that, for those skilled in the art, it is possible to make various improvements and modifications to the invention without departing from the principle of the invention, and those improvements and modifications also fall within the scope of the claims of the invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.