Household garbage AI intelligent unmanned transport vehicle and garbage treatment and transportation method

Abstract

The invention relates to the technical field of garbage disposal, and discloses a household garbage AI intelligent unmanned transport vehicle and a garbage disposal transport method.The transport vehicle comprises a garbage transport vehicle body, a data acquisition module, a path optimization module, a dynamic path adjustment module, a multi-vehicle collaborative scheduling module and an execution module; the method comprises the steps that a garbage transportation task is initialized, and garbage stations and the state of a transportation vehicle are collected; calculating an optimal path, and performing real-time adjustment based on dynamic data; by adopting multi-vehicle cooperative scheduling, path conflicts are avoided, and the transportation efficiency is optimized; garbage collection, transportation and delivery are completed, and transportation data are recorded. According to the method, efficient automation of garbage transportation is achieved through intelligent data collection, path optimization, dynamic adjustment and multi-vehicle cooperative scheduling, the optimal path is calculated through the variational method and the Euler-Lagrange equation, the driving route is dynamically adjusted in combination with the Hamilton-Jacobi-Bellman equation, multi-vehicle scheduling is optimized based on the differential game and the Nash equilibrium, and the efficiency of garbage transportation is improved. And high-efficiency transportation and no path conflict are ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of garbage treatment, and specifically to an AI intelligent driverless transport vehicle for domestic garbage and a garbage treatment and transportation method. Background Art

[0002] Currently, with the improvement of the intelligent level of urban management, the application of intelligent logistics systems is becoming increasingly widespread. Among them, intelligent driverless technology has shown efficient and convenient characteristics in aspects such as cargo transportation and distribution, effectively improving logistics efficiency and reducing labor costs.

[0003] At present, the collection and transportation of domestic garbage still mainly rely on manual driving. The labor intensity of drivers is high, their operating habits are not unified, and scheduling depends on experience, making it difficult to accurately match the needs of garbage stations. This leads to garbage overflow at some stations, over-frequent cleaning at some stations, increased transportation costs, and low efficiency.

[0004] Some cities have tried autonomous driving, but mostly use preset routes. The vehicle cannot adjust its driving path according to real-time traffic. In case of emergencies such as construction or accidents, it can only wait passively, resulting in transportation delays. At the same time, multiple garbage transport vehicles may repeatedly drive on the same section of the road, and the scheduling lacks optimization, increasing time and energy consumption.

[0005] In terms of path optimization, traditional Dijkstra or A* algorithms only consider the shortest distance and do not fully consider factors such as vehicle load, energy consumption, and road gradient. Some calculated shortest paths contain a large number of sharp turns or narrow sections, which are not suitable for garbage transport vehicles and may affect driving stability and increase fuel consumption. In addition, traditional path optimization methods are difficult to take into account multi-vehicle coordination, and path conflicts are likely to occur, affecting the overall transportation efficiency.

[0006] Multi-vehicle scheduling still mainly relies on static rules or simple polling, and fails to dynamically match the needs of garbage stations and traffic conditions. This may lead to garbage transport vehicles gathering for cleaning or concentrating in the same area during peak hours, exacerbating traffic pressure and further reducing efficiency.

[0007] In terms of driving control, traditional PID control has a slow response and is difficult to adapt to complex urban road conditions such as pedestrians, non-motor vehicles, intersections, etc., affecting driving smoothness and increasing energy consumption. In addition, existing garbage transport vehicles do not integrate an energy recovery system, and the frequent start-stop garbage collection and transportation mode results in waste of braking energy and increases operating costs.

[0008] Therefore, the present invention proposes an AI intelligent driverless transport vehicle for domestic garbage and a garbage treatment and transportation method to solve the deficiencies of the prior art. Summary of the Invention

[0009] In view of the deficiencies of the prior art, the present invention provides a domestic waste AI intelligent driverless transport vehicle and a waste treatment and transportation method, which solve the problems of high labor intensity in manual driving in the existing waste transportation mode, lack of real-time adjustment in fixed-path autonomous driving, low efficiency and easy conflict in multi-vehicle scheduling, unoptimized energy consumption and load in path planning, and poor adaptability in driving control.

[0010] To achieve the above objectives, the present invention is realized through the following technical solutions: A domestic waste AI intelligent driverless transport vehicle, comprising: A waste transport vehicle for performing waste collection and transportation tasks, which includes an on-vehicle power supply, a drive system, an autonomous driving control unit, and a communication module; A data acquisition module for collecting the waste volume, waste type, and location information of waste sites, obtaining the processing capacity and location information of waste treatment stations, and simultaneously obtaining the current location and load status of the waste transport vehicle; A path optimization module for calculating the initial optimal path of the waste transport vehicle based on the variational method and solving the optimal path using the Euler-Lagrange equation to obtain the optimal driving path of the transport vehicle; A dynamic path adjustment module for dynamically optimizing and adjusting the optimal driving path of the waste transport vehicle based on the Hamilton-Jacobi-Bellman equation in combination with real-time traffic data; A multi-vehicle collaborative scheduling module for optimizing the scheduling strategy of waste transport vehicles based on differential game theory in the case of multiple waste transport vehicles running simultaneously, and optimizing the optimal driving path of waste transport vehicles based on Nash equilibrium; An execution module for controlling the acceleration, braking, and steering of the waste transport vehicle according to the optimized path and completing the tasks of collecting and transporting waste to the target waste treatment station.

[0011] The present invention also provides a domestic waste treatment and transportation method, including the following steps: Initialization of waste transport tasks, obtaining the status information of waste sites, waste treatment stations, and waste transport vehicles; Path optimization calculation, calculating the initial optimal path of the waste transport vehicle based on the variational method and solving the Euler-Lagrange equation to obtain the optimal path; Dynamic path adjustment, dynamically optimizing and adjusting the driving path of the waste transport vehicle based on the Hamilton-Jacobi-Bellman equation in combination with real-time traffic data; Multi-vehicle collaborative scheduling, optimizing multi-vehicle transport scheduling based on differential game theory and using Nash equilibrium to solve the optimal driving path of each waste transport vehicle; Waste delivery and task completion, the waste transport vehicle drives to the target waste treatment station to unload the waste, and records the data of this transportation to optimize subsequent transportation tasks.

[0012] The present invention provides a domestic waste AI intelligent driverless transport vehicle and a waste treatment and transportation method. Having the following Beneficial effects: 1. The optimized design of the waste transport vehicle in the present invention breaks through the limitations of traditional manual driving and fixed preset route autonomous driving. By adopting an autonomous driving control unit, combined with dynamic path optimization, intelligent scheduling and adaptive control, fully automatic and efficient waste transportation is realized. Compared with the traditional manual driving mode, the present invention reduces the labor intensity and avoids the operation instability caused by human factors; compared with the fixed path autonomous driving scheme, the present invention can perceive the traffic conditions in real time and dynamically adjust the driving route, improving the transportation efficiency in complex environments.

[0013] 2. The present invention uses the variational method and the Euler-Lagrange equation to accurately calculate the optimal path of the waste transport vehicle. By optimizing the path curvature and reasonably distributing the speed, the driving is ensured to be smoother and more energy-efficient. Compared with the existing path planning based on Dijkstra or A* algorithms, unnecessary sharp turns and detours are reduced, the overall transportation efficiency is improved, and the problem that the traditional method is prone to local optimum rather than global optimum in complex road conditions is solved.

[0014] 3. Combined with the Hamilton-Jacobi-Bellman equation, the present invention can analyze the traffic conditions in real time and dynamically optimize the transportation path. With the help of traffic flow, signal light timing, and sudden congestion data, the system can automatically avoid peak sections. Compared with the traditional fixed path planning method, this scheme can effectively reduce the delays caused by traffic emergencies and solve the deficiencies of the existing technology such as rigid paths and lack of flexibility.

[0015] 4. Based on differential game theory, the present invention uses the Nash equilibrium to solve the optimal scheduling scheme. Multiple waste transport vehicles operate in coordination, with reasonable task allocation and avoidance of path conflicts. Compared with the traditional single vehicle independent scheduling method, this scheme ensures the maximization of the global transportation efficiency, reduces repeated transportation and resource waste, and solves the scheduling conflict problem existing in the multi-vehicle operation of the existing technology.

[0016] 5. By adopting model predictive control, the present invention can accurately calculate acceleration, braking and steering strategies, making the waste transport vehicle drive more smoothly, avoiding sudden stops and sharp turns, and improving safety. At the same time, combined with an intelligent perception system, the vehicle operation is adjusted in real time. Compared with the traditional PID control method, this scheme performs better in complex environments and solves the problems of lagging response and poor adaptability of the existing control methods. Description of the Drawings

[0017] Figure 1 is an architecture diagram of a domestic waste AI intelligent driverless transport vehicle of the present invention; Figure 2 Structural schematic diagram of the garbage transport vehicle of the present invention; Figure 3 Structural schematic diagram of the data acquisition module of the present invention; Figure 4 Structural schematic diagram of the multi-vehicle collaborative scheduling module of the present invention; Figure 5 Structural schematic diagram of the execution module of the present invention; Figure 6 Method flow chart of the present invention. Specific implementation manners

[0018] Next, in combination with the accompanying drawings of the specification of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0019] Please refer to Figures 1 - 5 , an AI intelligent driverless domestic garbage transport vehicle provided by an embodiment of the present invention includes: a garbage transport vehicle for performing garbage collection and transportation tasks, which includes an on-vehicle power supply, a drive system, an autonomous driving control unit, and a communication module; As the core entity for performing garbage collection and transportation tasks, the garbage transport vehicle undertakes functions such as power drive, autonomous driving control, data transmission, and task execution of the vehicle. The garbage transport vehicle not only needs to adapt to the complex urban road environment, but also needs to have efficient energy management, intelligent perception, path tracking, and garbage collection and transportation capabilities. Therefore, its design needs to be closely connected with path optimization, dynamic path adjustment, multi-vehicle collaborative scheduling, and the execution module to ensure that the vehicle completes the garbage transportation task under efficient, safe, and stable conditions.

[0020] In this embodiment, the garbage transport vehicle includes an on-vehicle power supply, a drive system, an autonomous driving control unit, and a communication module, and integrates multiple sensor units to realize real-time perception and data transmission of the vehicle operation state, surrounding environment, and garbage stations.

[0021] On-vehicle power supply. Generally, the energy supply method of the garbage transport vehicle affects its endurance and operation cost. In this embodiment, the on-vehicle power supply can adopt a lithium-ion battery pack or a hydrogen fuel cell. Among them, the lithium-ion battery pack is suitable for urban short-distance transportation, while the hydrogen fuel cell can adapt to longer-distance and high-load transportation tasks.

[0022] As an option, the in-vehicle power system can also integrate an energy recovery system. By using regenerative braking technology, the mechanical energy generated when the vehicle decelerates is converted into electrical energy and stored in the battery pack, thus improving the endurance capacity.

[0023] In a possible implementation, the garbage truck can be combined with an intelligent battery management system (BMS) to monitor the battery status in real time, including voltage, current, temperature, and remaining power. Through predictive algorithms, the charging and discharging strategies are optimized to extend the battery life.

[0024] Drive system. In this embodiment, the drive system includes a motor drive unit, a transmission system, and a chassis control unit, ensuring that the garbage truck has good power performance and handling stability in complex environments.

[0025] Specifically, the motor drive unit uses a permanent magnet synchronous motor (PMSM) or in-wheel motors. The former is suitable for high-efficiency driving, while the latter can improve space utilization and reduce power transmission losses.

[0026] In a possible implementation, the drive system of the garbage truck supports torque vectoring control (TVC). Under different load and road conditions, the torque distribution of the left and right wheels is dynamically adjusted to improve the vehicle's stability and steering ability.

[0027] As an option, the chassis control unit integrates an adaptive suspension system. Combining the load status of the garbage truck and the road conditions, the suspension stiffness and damping are adjusted in real time to enhance driving comfort and stability.

[0028] Autopilot control unit. In this embodiment, the autopilot control unit is responsible for the intelligent control of the driving path, speed, steering, and obstacle avoidance of the garbage truck. The core components include a high-performance computing platform, an environment perception module, and a path tracking control system.

[0029] Generally, the computing platform adopts a GPU+CPU heterogeneous architecture. Combining deep learning algorithms, it realizes object detection, path recognition, and trajectory optimization. In a possible implementation, the autopilot control unit uses neural network reinforcement learning (RL) to optimize the driving strategy based on historical driving data, improving path tracking accuracy and traffic adaptability.

[0030] Specifically, the path tracking control system uses model predictive control (MPC) to calculate the optimal acceleration, braking, and steering control strategies based on the dynamic model of the garbage truck. The dynamic model can be expressed as: Where, x is the lateral position of the vehicle (global coordinate system); y is the longitudinal position of the vehicle (global coordinate system); θ is the heading angle; v is the vehicle speed; a is the acceleration; δ is the steering angle; L is the wheelbase of the vehicle.

[0031] As an option, in path tracking control, the optimization objective function of the MPC controller can be defined as: Where, J is the optimization objective function, representing the overall error and control cost of the garbage truck in path tracking control; T is the prediction time domain window, that is, the number of future time instants considered for optimal control; x t is the state vector of the vehicle at time t; is the target state vector of the vehicle on the ideal path at time t; is the square of the two-norm of the state error, representing the deviation between the actual state and the target state; Q is the state error weight matrix, used to adjust the influence of the state deviation on the optimization objective; u t is the control input vector of the vehicle at time t, including the acceleration a t and the steering angle δ t ; ∥u t ∥ 2 is the square of the two-norm of the control input, representing the magnitude of the control quantity and avoiding violent operations; R is the control input weight matrix, used to adjust the influence of the control input and prevent excessive control changes.

[0032] In a possible implementation, the autonomous driving control unit can integrate an obstacle avoidance system based on lidar + millimeter-wave radar fusion perception, and dynamically adjust the driving strategy in combination with the calculation results of the path optimization module.

[0033] Communication module. In this embodiment, the communication module is used to realize real-time data interaction between the garbage truck and the cloud scheduling system, other transport vehicles, and road infrastructure.

[0034] Generally, the communication module supports 5G, V2X (Vehicle-to-Everything), and DSRC (Dedicated Short-Range Communications) technologies to ensure low latency and high reliability of data transmission.

[0035] As an option, the garbage truck can adopt a distributed data sharing architecture to upload the locally processed data to the cloud for use by the path optimization, dynamic path adjustment, and multi-vehicle collaborative scheduling modules, realizing information sharing and task collaboration among multiple garbage trucks.

[0036] In a possible implementation, the communication module integrates an edge computing unit to reduce the computing burden on the cloud and improve the response speed. The edge computing unit can preprocess real-time data, including: Compressing and storing the status data of garbage sites; Detecting anomalies in vehicle driving data; Pre-computing task scheduling information to improve scheduling efficiency.

[0037] A data acquisition module, which is used to collect the amount of garbage, garbage types, and location information of garbage sites, obtain the processing capacity and location information of garbage treatment stations, and at the same time obtain the current location and load status of garbage transport vehicles; The data acquisition module is the foundation of the entire domestic waste AI intelligent driverless transport vehicle system, responsible for real-time acquiring and processing data related to garbage transport vehicles and their surrounding environment. This data includes the amount, type, and location information of garbage at garbage sites, the processing capacity and location information of garbage treatment stations, as well as the current location and load status of the transport vehicle, etc. This information is the basis for modules such as path planning, dynamic adjustment, and collaborative scheduling. Therefore, the performance of the data acquisition module directly determines the overall accuracy and efficiency of the system.

[0038] The data acquisition module uses a variety of sensors and advanced data processing technologies to monitor and transmit key data in real time, providing the required data support for subsequent path optimization, dynamic adjustment, and multi-vehicle scheduling. Sensor unit.

[0039] Generally, the data acquisition module includes a multi-source sensor unit for real-time perception of the surrounding environment. The multi-source sensor unit consists of a lidar, a camera, an infrared sensor, and a GPS module. Specifically: Lidar: The lidar is mainly used to obtain the three-dimensional structure information of the environment. It generates a high-precision environmental map by emitting laser beams and receiving reflected signals, and can detect the shape, position, distance, and size of surrounding obstacles. Therefore, the lidar is very important for the perception of dynamic environments, especially in complex urban environments, where it can effectively avoid collisions, judge obstacles, and terrain information.

[0040] Camera: The camera is used to provide visual information, mainly for identifying road signs, traffic lights, traffic markings, and other visual elements related to driving safety. It can be combined with image processing algorithms to perform tasks such as object recognition and target tracking, providing auxiliary information for the path planning and dynamic adjustment of the transport vehicle.

[0041] Infrared sensor: The infrared sensor can be used to sense obstacles in low-light environments, enhancing the performance of the system at night or in bad weather. It is mainly used to detect the road conditions ahead, especially in short line-of-sight or low-light environments, which helps to improve safety.

[0042] GPS module: The GPS module provides global positioning system data, which is used to obtain the location information of the transport vehicle in real time and combine with other sensor data to ensure the precise positioning of the vehicle-mounted system.

[0043] In some embodiments, the data acquisition module further includes an edge computing unit. As the core component of data processing, the edge computing unit undertakes the task of real-time processing and analysis of data from multi-source sensors. The edge computing unit can immediately perform data preprocessing, screening, and preliminary calculations after receiving sensor data, thereby transmitting the processed data to the upper-layer module, reducing data transmission latency, and improving the system's response speed.

[0044] Specifically, the edge computing unit can process data in the following ways: Perform denoising, filtering, and interpolation processing on the raw data collected by the sensors to ensure the quality of the input data.

[0045] Fuse the data from different sensors to obtain a comprehensive environmental model. Through the fusion of sensor data, the environmental information in the current state can be deduced more accurately.

[0046] After processing, the data will be transmitted to the main control unit to provide basic data for modules such as path planning and dynamic path adjustment.

[0047] As an option, the edge computing unit can also run real-time algorithms as needed. For example, image recognition algorithms, obstacle detection algorithms, etc. can be run to further enhance the automation and intelligence of the system.

[0048] Specifically, the working process of the data acquisition module is as follows: Environmental data acquisition: Real-time environmental data is collected through sensors such as lidar, cameras, infrared sensors, and GPS modules. The sensors periodically scan and collect information about obstacles, road conditions, garbage sites, and other relevant information in the environment.

[0049] Real-time data processing: The edge computing unit performs preliminary processing on the collected raw data, including steps such as data denoising and filtering. The processed data is sent to the control unit and fused when necessary to obtain more accurate and comprehensive environmental information.

[0050] Data transmission and feedback: The processed data will be transmitted to modules such as path planning, dynamic path adjustment, and collaborative scheduling. These modules rely on sensor data to make decisions to adjust the driving path and optimize the transportation task.

[0051] Abnormal Data Detection and Feedback: The edge computing unit also has the function of abnormal data detection. When abnormal sensors or inconsistent data are detected, the system will automatically issue an alarm and supplement and correct the data according to the set redundancy mechanism to ensure the stability of the system.

[0052] As the first link of the system, the output data of the data acquisition module is crucial for all subsequent modules. Specifically: During the path optimization process, the real-time location, road conditions, and traffic information provided by the data acquisition module will be used as the input for the path planning module to ensure that the calculated optimal path meets the requirements of the actual environment.

[0053] During the dynamic path adjustment phase, the input of real-time traffic flow, signal timing, weather, etc. data makes the adjusted path more in line with the current actual traffic state, thereby reducing transportation time and energy consumption.

[0054] In the multi-vehicle collaborative scheduling module, the vehicle status and surrounding environment information provided by the data acquisition module will help the task allocation module determine the optimal scheduling plan and make timely adjustments according to the requirements of the path coordination module to ensure the coordination and efficiency of the scheduling of each transport vehicle.

[0055] The path optimization module is used to calculate the initial optimal path of the garbage transport vehicle based on the variational method and solve the optimal path using the Euler-Lagrange equation to obtain the optimal driving path of the transport vehicle; In this embodiment, the path optimization module is a key component of the system. Its main function is to calculate the optimal driving path for the garbage transport vehicle based on advanced optimization algorithms. Path optimization not only needs to consider real-time environmental data such as traffic flow and road conditions, but also needs to be adapted according to the actual situation of the vehicle to minimize transportation time, energy consumption, and improve driving stability. The output of this module will directly affect the effects of dynamic path adjustment and multi-vehicle collaborative scheduling. Therefore, the accuracy and efficiency of path optimization are crucial for the performance of the entire system.

[0056] The task of the path optimization module is to calculate the optimal path of the garbage transport vehicle through the variational method and the Euler-Lagrange equation. This process requires comprehensive calculations based on the current traffic conditions, road information, and the position, load, etc. of the transport vehicle, and finally obtains a path that meets both the shortest time and the most energy-saving requirements. During the optimization process, the smoothness of the path, compliance with traffic rules, and energy consumption control are all important factors that must be considered.

[0057] In this embodiment, the path optimization module mainly relies on the variational method and the Euler-Lagrange equation to calculate the optimal path. The variational method is a common optimization method widely used in the calculation of the shortest path and minimization of cost. In the path planning problem, the variational method is used to minimize the driving time, energy consumption, etc. between the starting point and the target point.

[0058] In this embodiment, the path optimization module mainly relies on the variational method and the Euler-Lagrange equation to calculate the optimal path. The variational method is a common optimization method widely used in the calculation of the shortest path and cost minimization. In path planning problems, the variational method is used to minimize the travel time, energy consumption, etc. between the starting point and the target point.

[0059] Specifically, based on the state variables and control inputs, the variational method defines a path cost function J(x,u), which represents the total cost from the current state x to the target state. The optimization objective of the cost function is to minimize the travel time, energy consumption, traffic cost, etc., and can usually be expressed as: where J(x,u) is the total cost function, representing the cumulative cost from the starting point to the target point, covering all costs incurred during the driving process, including time, energy consumption, traffic expenses, etc.; x represents the state of the system, such as the position, speed, load, etc. of the vehicle; u represents the control input, such as acceleration, steering wheel angle, etc.; L(x,u) is the instantaneous cost function, representing the cost of the system at a certain moment; h(x,u) is the future cost function, representing the cumulative cost in the future period; λ is the discount factor, controlling the impact of future costs on the current decision; T is the time termination point, representing the time range of path planning.

[0060] By minimizing this cost function, the control input sequence of the optimal path can be obtained, thus ensuring the optimal driving strategy.

[0061] In path optimization, the Euler-Lagrange equation is used to describe the motion equation of the system and the conditions for optimizing the path. Specifically, the Euler-Lagrange equation is used to determine the optimal curvature and optimal speed distribution of the path. By transforming the path optimization problem into the problem of solving the Euler-Lagrange equation, the path equation that meets the optimal control criterion can be obtained. The general form of the Euler-Lagrange equation is as follows: where represents the derivative with respect to time t, usually representing the process of changing with time; x is the state variable of the vehicle, representing a certain physical quantity or parameter of the system, such as the position, speed, acceleration, etc. of the vehicle; is the derivative of x, representing the speed of the vehicle or the dynamic behavior of the system; L is the Lagrangian, representing the cost function or Lagrangian of the system, usually the difference between kinetic energy and potential energy, or the cost function of the path; is the partial derivative of the Lagrangian with respect to the speed (or state variable )), usually related to kinetic energy, representing the impact of speed on the system cost; To represent the partial derivative of the Lagrangian with respect to the state variable x, it is usually related to the potential energy or other position-related cost terms, indicating the impact of position on the system cost.

[0062] By solving this equation, the system can obtain the optimal control strategy and path, enabling the vehicle to achieve the optimal driving trajectory in a changing environment.

[0063] Generally, the path optimization module will work according to the following steps: Initial path calculation: When the system starts, the path optimization module first calculates the preliminary optimal path from the starting position to the target processing station. This process relies on the variational method and the Euler-Lagrange equation, and the preliminary optimal path takes into account the current traffic state, road information, etc.

[0064] Path cost function construction: The system defines the path cost function L(x,u) and the future cost function h(x,u) according to the actual road conditions, and constructs a comprehensive cost function in combination with the discount factor λ and the time termination point T. This cost function comprehensively considers factors such as travel time, traffic flow, road conditions, energy consumption, etc., and calculates the optimal path.

[0065] Euler-Lagrange equation solution: After the initial path calculation, the path optimization module will optimize the details of the path by solving the Euler-Lagrange equation. The solution of the equation will ensure that the path reaches the optimal curvature, speed distribution, and driving stability without violating traffic rules.

[0066] Path feasibility verification: After the path calculation is completed, the system will perform path feasibility verification. This step takes into account factors such as traffic signals, road surface conditions, slopes, etc., to ensure that the calculated path is feasible under actual traffic conditions. If some sections are not feasible, the system will re-plan the path and make adjustments.

[0067] In a possible implementation, the path optimization module is closely connected to the data acquisition module, the dynamic path adjustment module, and the multi-vehicle cooperative scheduling module. Specifically: Connection with the data acquisition module: The data acquisition module provides real-time information such as traffic flow, road conditions, and weather. This information provides key inputs for the path optimization module. The path optimization module dynamically adjusts the path according to this information to ensure that the path planning can adapt to the current traffic conditions.

[0068] Connection with the dynamic path adjustment module: The preliminary optimal path calculated by the path optimization module provides the basic input for the dynamic path adjustment module. The dynamic path adjustment module fine-tunes the path based on real-time monitoring of traffic conditions and according to real-time data (such as traffic jams, accidents, etc.) to ensure that the transport vehicle is always driving on the optimal route.

[0069] Connected to the multi-vehicle collaborative scheduling module: Based on the optimal path provided by the path optimization module, the multi-vehicle collaborative scheduling module further coordinates among multiple vehicles to ensure that there are no path conflicts among vehicles within the same or adjacent time periods. The scheduling system optimizes the paths of multiple vehicles through the Nash equilibrium algorithm to ensure the globally optimal transportation effect.

[0070] The dynamic path adjustment module is used to dynamically optimize and adjust the optimal driving path of the garbage truck based on the Hamilton-Jacobi-Bellman equation combined with real-time traffic data; In this embodiment, the main task of the dynamic path adjustment module is to dynamically optimize and adjust the optimal driving path of the garbage truck based on the Hamilton-Jacobi-Bellman (HJB) equation combined with real-time traffic data. The dynamic path adjustment module continuously receives and processes real-time traffic information from sources such as traffic monitoring systems and data acquisition modules, and dynamically adjusts the calculated optimal path to cope with sudden traffic condition changes, such as traffic jams and accidents. Through this module, the system can ensure that the garbage truck always drives on the optimal path, maximizing transportation efficiency, reducing delays, and achieving flexible adjustment in a changing traffic environment.

[0071] This module not only optimizes the preliminary path but also can correct the path according to real-time situations to avoid traffic congestion or other factors affecting transportation efficiency. Through the application of the HJB equation, the system can provide optimal decisions for path selection at each moment to ensure that the vehicle can reach the destination quickly and safely.

[0072] The Hamilton-Jacobi-Bellman equation (HJB equation) is one of the basic tools in optimal control theory and is used to describe the optimal decision-making process of a dynamic optimization problem. By solving the HJB equation, the system can provide the optimal control strategy for each state, thereby calculating the optimal path.

[0073] In this embodiment, the form of the HJB equation is: Among them, V * (x) is the optimal value function, which represents the lowest cost or maximum benefit that can be obtained starting from the current state x according to the optimal strategy; u is the control variable, referring to the controllable input of the system; g(x,u) is the immediate cost, which represents the immediate cost when the system is in state x and the control input u is applied at a certain moment; is the discounted cumulative future cost; h(x,u) represents the future cost under state x and control input u; e -λtis the discount factor; λ is the discount rate; T is the time termination point, representing the time range of the optimization problem; λ is the discount factor, which controls the impact degree of future costs.

[0074] In this embodiment, the dynamic path adjustment module performs path optimization according to real-time traffic data and the current position of the vehicle through the following steps: Real-time traffic data collection and update: The data collection module continuously monitors and collects real-time traffic-related data, such as road traffic flow, traffic accidents, construction conditions, etc. This data is transmitted to the dynamic path adjustment module through the communication module.

[0075] Initial path calculation: In the initial stage, the path optimization module has calculated the initial optimal path according to the variational method and the Euler-Lagrange equation. However, in practical applications, traffic conditions change continuously, so the system needs to dynamically adjust the path to adapt to the new traffic data.

[0076] Solution of the HJB equation: The dynamic path adjustment module uses real-time traffic information and takes the HJB equation as an optimization tool for solution. At each moment, the module calculates the optimal control strategy through the HJB equation to obtain the optimal path adjustment at the corresponding time and location. For example, when it is found that a certain section is blocked, the system will calculate a new path according to the real-time data to ensure that the transport vehicle will not be delayed.

[0077] Path adjustment and optimization: According to the solution result of the HJB equation, the dynamic path adjustment module optimizes and adjusts the existing path. If the traffic condition changes, the system will immediately calculate a new path and instruct the vehicle to drive along the new optimal route. Specifically, the system adjusts the control input u (such as acceleration, steering angle, etc.) so that the vehicle can quickly switch to the new path.

[0078] Feedback and correction: To ensure the real-time adaptation of the optimal path, the path adjustment module continuously monitors the running state of the vehicle and continuously feedbacks and updates the path according to the current vehicle position and traffic data. Through this closed-loop control, the system can ensure that the transport vehicle drives on the optimal route at each moment.

[0079] The cooperation between the dynamic path adjustment module and other modules is the key to the efficient operation of the system. In this embodiment, the path adjustment module is closely connected with the data collection module, the path optimization module, and the multi-vehicle scheduling module: Connection with the data collection module: The data collection module provides the latest traffic flow, road conditions and other information in real time, providing the necessary input for the dynamic path adjustment module. Through the seamless connection with the data collection module, the dynamic path adjustment module can obtain the changing traffic conditions in the first time.

[0080] Connected to the path optimization module: The optimal path initially calculated by the path optimization module provides an initial reference for the dynamic path adjustment module. The dynamic path adjustment module adjusts and optimizes the path based on real-time traffic data.

[0081] Connected to the multi-vehicle scheduling module: The multi-vehicle scheduling module coordinates according to the path planning and adjustment results of multiple transport vehicles to avoid path conflicts between different vehicles. Through collaborative work, multiple vehicles can share the optimal path and jointly adjust to achieve the global optimum.

[0082] A multi-vehicle collaborative scheduling module, which is used to optimize the scheduling strategy of garbage transport vehicles when multiple garbage transport vehicles are running simultaneously based on differential game theory, and optimize the optimal driving path of garbage transport vehicles based on Nash equilibrium; In this embodiment, the multi-vehicle collaborative scheduling module is used to optimize the scheduling strategy of garbage transport vehicles based on differential game theory and calculate the optimal driving paths of multiple garbage transport vehicles using Nash equilibrium. When multiple vehicles are running simultaneously, the optimization of scheduling needs to consider multiple factors, including the mutual influence between vehicles, resource competition on the path, reasonable allocation of transportation tasks, etc. The goal of this module is to ensure the maximization of the overall transportation efficiency of the system while avoiding path conflicts or unnecessary resource occupation between individual transport vehicles.

[0083] Generally, this module is closely connected to the dynamic path adjustment module and the path optimization module. The path optimization module first calculates the initial optimal path for each vehicle, and the dynamic path adjustment module makes real-time corrections to the path and provides the adjusted path data to the multi-vehicle collaborative scheduling module. In a possible implementation, the multi-vehicle collaborative scheduling module performs global optimization based on this data using differential game theory, thereby achieving reasonable scheduling among vehicles at the system level.

[0084] In this embodiment, the multi-vehicle collaborative scheduling module includes the following four sub-modules, each undertaking different optimization calculation tasks to ensure the optimal scheduling effect at the global level: Task allocation module: Generally, the task allocation module is used to reasonably allocate transportation tasks based on the demand situation of garbage sites and the status information of each garbage transport vehicle to ensure the balance and global optimality of the tasks.

[0085] As an option, this module first collects real-time demand data of garbage sites, including the filling degree of garbage bins, cleaning frequency, etc. Then, combined with the status of each vehicle (including current load, position, driving speed, remaining fuel, etc.), an optimal allocation algorithm is used to allocate tasks.

[0086] In a possible implementation, the task allocation module uses the linear programming method to optimize the task allocation objective, and the objective function is as follows: where \(x_{ij}\) ij indicates whether the garbage truck \(i\) is assigned to the garbage site \(j\), and the value is 0 or 1; \(c_{ij}\) ij represents the transportation cost of the garbage truck \(i\) going to the garbage site \(j\), taking into account factors such as driving distance, time, fuel consumption, etc.; \(N\) is the total number of garbage trucks, and \(M\) is the total number of garbage sites.

[0087] Specifically, the task of this module is to solve the optimal \(x_{ij}\) ij combination to minimize the global transportation cost while meeting the transportation requirements of all garbage sites.

[0088] Path coordination module: As an option, the path coordination module is used to calculate the driving paths of multiple garbage trucks, ensure that the paths do not conflict, and optimize the global traffic flow.

[0089] Generally, this module depends on the conflict detection and path correction algorithm. After the preliminary path calculation of the garbage trucks, it detects whether there are potential conflict problems such as path intersections and blockages for all vehicle paths. If a conflict is found, this module uses a heuristic path adjustment method to appropriately adjust the paths of some vehicles.

[0090] In a possible implementation, this module adopts a path optimization strategy based on the Lagrangian relaxation method to construct a constrained optimization problem: where \(x_i\) i is the state of the garbage truck \(i\); \(u_i\) i is its control input; \(x_j\) -i represents the states of other garbage trucks; \(L(x_i,u_i,x_j)\) i (\(x_i\) i , \(u_i\) i , \(x_j\) -i ) is the immediate cost function, including fuel consumption, driving time, congestion loss, etc.

[0091] Specifically, this module makes the driving paths of all garbage trucks globally optimal through optimization and solution, while avoiding path conflicts.

[0092] Revenue evaluation module: Generally, the revenue evaluation module is used to calculate the revenue function of each garbage truck based on transportation time, energy consumption, and congestion cost, and optimize the path strategy.

[0093] In a possible implementation, this module defines the revenue function of each garbage truck as follows: \(R_i\) i = \(\alpha_1T_i\) i + \(\alpha_2E_i\) i + \(\alpha_3C_i\) i; Among them, R i is the total revenue of garbage truck i; T i is the transportation time of garbage truck i; E i is the energy consumption (such as fuel consumption or power consumption); C i is the congestion cost, representing the additional time loss caused by traffic flow; α1, α2, α3 are weight parameters, representing the influence degree of different factors on the revenue.

[0094] Specifically, this module adjusts the path optimization strategy to ensure the rationality of global scheduling while maximizing the revenue function of each vehicle.

[0095] Nash equilibrium solving module: In a possible implementation, the Nash equilibrium solving module is used to optimize the optimal driving path of each garbage truck based on the Nash equilibrium and make the solution converge to a stable state through iterative optimization.

[0096] Generally, this module uses an algorithm based on Best-Response Dynamics to calculate the Nash equilibrium point. This algorithm iteratively adjusts the driving strategies of each vehicle so that no garbage truck can obtain a better result by unilaterally changing its own path.

[0097] Specifically, the system calculates the optimal control strategy of garbage truck i in the current state: Among them, is the optimal control input of garbage truck i, including decision variables such as acceleration, direction, and speed adjustment; u i is the control input variable of garbage truck i, and its goal is to find the optimal to minimize the cost function; J i (x i , u i , x -i ) is the cost function of garbage truck i, measuring the optimization goal of its transportation path; x i is the state variable of garbage truck i, including parameters such as its current position, speed, and load; x -i is the state set of all other garbage trucks except garbage truck i, representing the influence of other vehicles in the system; represents finding the optimal solution of u i so that the cost function J i (x i , u i , x -i ) reaches the minimum; during the calculation process, the strategies x -i of other vehiclesis considered fixed; through multiple iterative calculations, the strategies of all vehicles reach the Nash equilibrium point.

[0098] In one possible implementation, this module is combined with reinforcement learning methods, such as deep Q-network (DQN) or proximal policy optimization (PPO), to improve the efficiency of Nash equilibrium solving and accelerate the convergence of the algorithm.

[0099] In one possible implementation, this module works closely with other path optimization and adjustment modules to achieve global optimal scheduling: Connection with the route optimization module: The task allocation module uses the initial route data provided by the route optimization module to allocate and adjust the transportation tasks; Connection with the dynamic path adjustment module: The path coordination module receives the real-time optimized path provided by the dynamic path adjustment module and corrects the path conflicts; Connection with the data acquisition module: The revenue evaluation module and the Nash equilibrium solution module rely on the real-time traffic data of the data acquisition module to optimize transportation revenue and adjust the equilibrium calculation strategy.

[0100] The execution module is used to control the acceleration, braking and steering of the garbage truck according to the optimized path, and complete the task of collecting and transporting the garbage to the target garbage treatment station; In this embodiment, the execution module is a key component of the AI ​​intelligent unmanned domestic garbage transport vehicle. Its main task is to accurately control the acceleration, braking and steering of the garbage transport vehicle according to the optimized path to ensure that the vehicle can efficiently and safely complete the garbage collection, transportation and delivery tasks. To achieve this goal, the execution module adopts a method based on model predictive control (MPC) and combines intelligent perception with dynamic adjustment strategies to accurately control the dynamic behavior of the garbage transport vehicle.

[0101] The path execution unit controls the driving state of the garbage truck according to the optimal path information provided by the path optimization module. The path execution unit uses the model predictive control (MPC) method, combined with the vehicle's dynamic model, to accurately calculate the acceleration, braking and steering control strategies.

[0102] Under this control method, MPC first predicts the behavior of the vehicle in the future based on the current vehicle state and the predetermined target path, and optimizes the control input based on the predicted results. The control goal is usually to make the vehicle travel along the optimized path as much as possible, avoid collisions with other vehicles or obstacles, and meet the vehicle's dynamic constraints.

[0103] Specifically, the dynamic equation of the vehicle can be expressed as: in, For the vehicle state vector x t is the time derivative, representing the rate of change of the vehicle's state at time t. x t is the state vector of the vehicle, usually including the vehicle's position information, heading angle, and speed; u t is the control input vector of the vehicle, usually including acceleration and steering angle; f(x t , u t ) is the vehicle dynamics model, which describes the law of change of the vehicle state over time given the state x t and the control input u t .

[0104] In MPC, the goal is to minimize the cost function through the control input u t , and the cost function can be expressed as: where is the predefined target state (position, speed, and heading angle) on the optimized path, usually provided by the path planning module; Q and R are the weighting matrices of the state and control input respectively, used to balance the accuracy of path tracking and the smoothness of the control input; represents the deviation cost between the vehicle and the target path, aiming to keep the vehicle as close as possible to the optimal path; ∥u t ∥ 2 represents the cost of the control input, aiming to prevent excessive acceleration or steering and make the vehicle operate smoothly.

[0105] The optimization process of MPC includes the following steps: State prediction: Based on the current state x t and the control input u t , predict the state x t+k of the vehicle within a future period of time (where k is the prediction step).

[0106] Optimization objective: Define a cost function J, which is optimized according to the deviation between the current state and the target state, as well as the cost of the control input.

[0107] Constraint conditions: Consider the dynamic constraints of the vehicle, including limitations such as maximum acceleration, maximum braking force, and minimum turning radius. For example, the maximum acceleration a max and the minimum turning radius R min constraints can be expressed as: a min ≤a t ≤a max , |δ t | ≤ δ max ; where, a tThe acceleration control input for the vehicle at time t, representing the acceleration of the vehicle at that moment; a min The minimum value of the vehicle acceleration, restricting the minimum acceleration of the vehicle, usually representing the maximum deceleration (i.e., maximum braking) that the vehicle can exert; a max The maximum value of the vehicle acceleration, restricting the maximum acceleration of the vehicle, usually representing the maximum acceleration ability that the vehicle can exert; δ t The steering angle control input for the vehicle at time t, representing the angle of the vehicle's front wheels; δ max The maximum limit value of the vehicle steering angle, restricting the maximum amplitude of the front wheel steering angle. This constraint is to avoid the vehicle turning too sharply, resulting in unstable driving.

[0108] Optimization solution: Use numerical optimization methods (such as QP optimization or gradient descent method) to solve the optimization problem, so as to obtain the optimal control input sequence within a future period of time Control input update: According to the optimization result, select the optimal control input at the current time t and perform acceleration, braking or steering control.

[0109] Garbage collection unit: The garbage collection unit is responsible for loading garbage at the garbage site and recording the weight and type of the loaded garbage. In this embodiment, the garbage collection unit combines intelligent sensing technology and weighing sensors to detect the status of the garbage container in real time, including information such as the weight, type, and filling degree of the garbage.

[0110] The garbage collection unit monitors and records the loading situation of the garbage container in real time through sensors installed on the vehicle body. Each time it loads, the garbage collection unit records the weight information W of the garbage load , and determines the type S of the garbage through the garbage classification sensor type . This information will be used for subsequent path optimization and task scheduling, so as to perform more efficient task allocation when dealing with different types of garbage. The loading weight W of the garbage load can be expressed by the following formula: where, w i is the weight of each garbage unit, and n is the total number of garbage units.

[0111] Garbage delivery unit: The garbage delivery unit is responsible for unloading garbage at the garbage treatment station and updating the garbage status information of the garbage site and the treatment station. When the garbage transport vehicle arrives at the target garbage treatment station, the delivery unit automatically unloads the garbage according to the capacity and status information of the garbage treatment station and updates the garbage treatment information of the site.

[0112] The delivery unit detects the capacity and status of the waste station through sensing devices to ensure the smooth progress of the unloading process. Specifically, the delivery unit conducts data interaction with the management system of the target waste station to update information such as the type of waste, the processing status, and the remaining capacity in real time.

[0113] For example, the weight W of the waste after unloading unload and the waste processing status S status can be updated in the following ways: where, w j is the weight of each unloading unit, and m is the total number of unloading units.

[0114] During the driving process of the waste transport vehicle, the execution module not only needs to ensure that the vehicle travels along the optimized path, but also needs to respond in real time to emergencies (such as road obstacles, traffic congestion, etc.). In this case, the execution module can combine the Dynamic Window Approach (DWA) and path tracking control technology to dynamically adjust the vehicle path according to environmental changes to avoid collisions and traffic conflicts.

[0115] In some embodiments, the execution module is also responsible for monitoring the energy consumption of the vehicle. In the case of long - distance driving or heavy load, the execution module will reduce the energy consumption by optimizing the control strategy to extend the vehicle's endurance time.

[0116] Please refer to Figure 6 , the present invention also provides a method for treating and transporting domestic waste, including the following steps: S1. Initialize the waste transport task to obtain the status information of the waste station, the waste treatment station, and the waste transport vehicle; S2. Optimize the path calculation, calculate the initial optimal path of the waste transport vehicle based on the variational method, and solve the Euler - Lagrange equation to obtain the optimal path; S3. Dynamically adjust the path, dynamically optimize and adjust the driving path of the waste transport vehicle based on the Hamilton - Jacobi - Bellman equation combined with real - time traffic data; S4. Coordinate the multi - vehicle scheduling, optimize the multi - vehicle transport scheduling based on differential game theory, and use the Nash equilibrium to solve the optimal driving path of each waste transport vehicle; S5. Deliver the waste and complete the task, the waste transport vehicle travels to the target waste treatment station to unload the waste, and record the data of this transportation to optimize the subsequent transportation tasks.

[0117] For S1, in the task initialization phase, the system needs to obtain and integrate multiple data sources to ensure the efficient execution of garbage transportation tasks. Specifically, it includes the following: garbage site information collection, where the system automatically obtains data such as the geographical location, garbage stock, garbage classification information (such as recyclable waste, kitchen waste, hazardous waste, etc.), the overflow level of trash cans, and historical collection frequencies of each garbage site; garbage treatment station information collection, recording information such as the treatment capacity, remaining treatment capacity, and garbage classification treatment priorities of each garbage treatment station to ensure that garbage transport vehicles are matched with sites according to the optimal treatment capacity; obtaining the status of garbage transport vehicles, where the system detects information such as the number of currently available garbage transport vehicles, load conditions, driving status, power or fuel status (for new energy vehicles), and maintenance conditions to reasonably allocate transportation tasks; task allocation, where based on the demands of garbage sites and the receiving capacity of garbage treatment stations, the system initially allocates transportation tasks and generates an initial scheduling plan.

[0118] For S2, to ensure that garbage transport vehicles can complete tasks efficiently, the system uses path optimization methods to calculate the optimal driving path, mainly including: path initialization, based on the geographical distribution of garbage sites, garbage treatment stations, and transport vehicles, initially setting the driving path to ensure that the path covers all garbage sites that need to be served and can reach the garbage treatment station smoothly; optimizing path calculation, using the variational method to optimize the path of garbage transport vehicles, considering factors such as road conditions, traffic lights, driving time, energy consumption, and garbage loading volume to reduce transportation time and energy consumption and improve the overall transportation efficiency; path feasibility analysis, verifying the feasibility of the calculated path to ensure that the transport vehicle can pass smoothly and avoid narrow roads, height and weight limits, or traffic control sections.

[0119] For S3, during the driving process of garbage transport vehicles, the system monitors road conditions in real time and dynamically adjusts the path according to traffic data to ensure transportation efficiency. This process includes: obtaining real-time traffic data, obtaining the latest road information from traffic management systems, GPS data, road sensors, etc., including factors such as traffic congestion, accidents, and construction; recalculating the path, based on the current driving path and the latest traffic conditions, and combining with the path optimization algorithm, dynamically adjusting the driving path of the garbage transport vehicle to ensure the optimal driving plan; coping with emergencies, when emergencies occur (such as road closures, accidents, bad weather, etc.), the system will immediately re-plan a feasible path for the garbage transport vehicle or direct it to the nearest alternative garbage treatment station; vehicle autonomous decision-making, where the garbage transport vehicle can make autonomous decisions based on its own sensor data to determine whether to adjust the driving route to improve the degree of intelligence.

[0120] For S4, when multiple garbage transport vehicles are performing tasks simultaneously, the system will optimize the scheduling of multiple vehicles to improve the overall transportation efficiency and reduce unnecessary repeated transportation, specifically including: Task allocation optimization: Optimize the task allocation of multiple vehicles based on the load capacity, driving route, and real-time location of garbage transport vehicles to avoid duplicate transportation or resource waste; Collaborative path planning: Ensure that multiple garbage transport vehicles operate collaboratively at different times and on different routes, reduce congestion and unnecessary waiting time, and improve the efficiency of garbage collection and transportation; Dynamic task adjustment: When a transport vehicle fails to complete a task on time due to a breakdown, traffic jam, or other reasons, the system will reallocate the task to ensure that the garbage treatment process is not interrupted; Balanced load: Consider the load conditions of each garbage treatment station, reasonably allocate the garbage delivery volume of each transport vehicle, and avoid overloading or insufficient processing capacity of individual garbage treatment stations.

[0121] For S5, after the garbage transport vehicle arrives at the target garbage treatment station, it completes the garbage unloading and records the transport data to optimize subsequent transport tasks: Garbage delivery operation: According to the guidance of the garbage treatment station, the garbage transport vehicle unloads the garbage according to the classification requirements to ensure that different types of garbage enter the correct treatment process; Data recording and feedback: The system records data such as the driving trajectory, transport time, garbage loading volume, and unloading location of the transport vehicle and stores it in the database to optimize future transport task scheduling; Vehicle status inspection: After the garbage is unloaded, the system will check the vehicle status, including remaining power, fuel, vehicle condition, etc., to decide whether to carry out the next transport task or guide the vehicle back to the charging station or repair station; Transport summary and optimization: Based on the collected data, the system analyzes the efficiency of this transport task and provides optimization suggestions, such as improving route planning, adjusting the collection frequency of garbage stations, etc., to improve the overall garbage treatment and transportation.

[0122] Although the embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An AI intelligent driverless transport vehicle for domestic waste, characterized in that, Including: A garbage transport vehicle, which is used to perform garbage collection and transportation tasks, and includes an on-vehicle power supply, a drive system, an autonomous driving control unit, and a communication module; A data acquisition module, which is used to collect the garbage volume, garbage type, and location information of garbage sites, obtain the processing capacity and location information of garbage treatment stations, and at the same time obtain the current position and load status of the garbage transport vehicle; A path optimization module, which is used to calculate the initial optimal path of the garbage transport vehicle based on the variational method and use the Euler-Lagrange equation to solve the optimal path to obtain the optimal driving path of the transport vehicle; A dynamic path adjustment module, which is used to dynamically optimize and adjust the optimal driving path of the garbage transport vehicle based on the Hamilton-Jacobi-Bellman equation combined with real-time traffic data; A multi-vehicle collaborative scheduling module, which is used to optimize the scheduling strategy of garbage transport vehicles based on differential game theory when multiple garbage transport vehicles are running simultaneously, and optimize the optimal driving path of garbage transport vehicles based on Nash equilibrium; An execution module, which is used to control the acceleration, braking, and steering of the garbage transport vehicle according to the optimized path and complete the tasks of garbage collection and transportation to the target garbage treatment station.

2. The AI intelligent driverless transport vehicle for domestic waste according to claim 1, characterized in that, The data acquisition module includes: A multi-source sensor unit, including a lidar, a camera, an infrared sensor, and a GPS module, which is used to collect garbage site and road environment data in real time; An edge computing unit, which is used to process the collected garbage volume, type, and location data in real time.

3. An AI intelligent driverless transport vehicle for domestic waste according to claim 1, characterized in that, The optimal driving path includes the following: Path curve optimization, based on the garbage site location, the starting position of the garbage transport vehicle, and the target garbage treatment station, optimize the path curvature to reduce unnecessary sharp turns and detours; Speed distribution optimization, reasonably distribute the speed of the transport vehicle according to the road speed limit, intersections, slopes, and traffic conditions, reduce energy consumption, and improve driving stability; Optimal energy consumption control, considering the load, acceleration, and braking processes of the garbage transport vehicle, optimize the energy consumption model to reduce the total energy consumption during driving; Path feasibility verification, based on road traffic rules, traffic light timings, and real-time traffic information, verify whether the calculated driving path meets the actual traffic conditions and adjust infeasible sections.

4. The AI intelligent driverless transport vehicle for domestic waste according to claim 1, characterized in that, The steps of dynamically optimizing and adjusting the optimal driving path of the garbage transport vehicle based on the Hamilton-Jacobi-Bellman equation combined with real-time traffic data include: Construct a path cost function, with the minimum of driving time, energy consumption, and traffic flow as the optimization goal, and establish an optimal control problem in combination with the state variables of the garbage transport vehicle; Establish the Hamilton-Jacobi-Bellman equation, based on the path cost function and the dynamic model of the garbage transport vehicle, construct a state evolution equation to describe the optimal values of different states during the path adjustment process; Real-time data input, collect traffic flow, traffic light timings, road congestion conditions, and weather information, update the path state parameters; solve the optimal path adjustment strategy, calculate the optimal control strategy under different path states based on the dynamic programming method, and generate a new driving path; Path execution and feedback, the optimized path is sent to the execution module and corrected in real time according to the actual driving state of the transport vehicle.

5. An AI intelligent driverless transport vehicle for domestic waste according to claim 4, characterized in that, The Hamilton-Jacobi-Bellman equation is as follows: Among them, V * (x) is the optimal value function, representing the lowest cost or the maximum benefit that can be obtained starting from the current state x according to the optimal strategy; u is the control variable, referring to the controllable input of the system; g(x, u) is the immediate cost, representing the immediate cost at a certain moment when the system is in state x and the control input u is applied; is the discounted cumulative of future costs; h(x, u) represents the future cost under state x and control input u; e -λt is the discount factor; λ is the discount rate; T is the time termination point, representing the time range of the optimization problem; λ is the discount factor, controlling the influence degree of future costs.

6. The AI intelligent driverless transport vehicle for domestic waste according to claim 1, wherein, The multi-vehicle collaborative scheduling module includes: A task allocation module for allocating the optimal transportation task based on the waste site requirements and the state of the transport vehicle; A path coordination module for calculating the driving paths of multiple waste transport vehicles to avoid path conflicts; A revenue evaluation module for calculating the revenue function of each waste transport vehicle based on transportation time, energy consumption, and congestion costs, and optimizing the path strategy; A Nash equilibrium solving module for optimizing the optimal driving path of each waste transport vehicle based on Nash equilibrium and iteratively optimizing until convergence.

7. An AI intelligent driverless transport vehicle for domestic waste according to claim 1, characterized in that, The execution module includes: A path execution unit for controlling the acceleration, braking, and steering of the waste transport vehicle according to the optimized path; A waste collection unit for loading waste at the waste site and recording the loading weight and waste type; a waste delivery unit for unloading waste at the waste treatment station and updating the waste status information of the waste site and the treatment station.

8. An AI intelligent driverless transport vehicle for domestic waste according to claim 1, characterized in that, The execution module calculates the optimal acceleration, braking, and steering control strategies of the waste transport vehicle based on the model predictive control method.

9. A method for treating and transporting domestic waste, which is applied to an AI intelligent driverless transport vehicle for domestic waste as described in any one of claims 1-8, and is characterized in that, It includes the following steps: Initialization of waste transportation tasks, obtaining the status information of waste sites, waste treatment stations, and waste transport vehicles; Path optimization calculation, calculating the initial optimal path of the waste transport vehicle based on the variational method and solving the Euler-Lagrange equation to obtain the optimal path; Dynamic path adjustment, dynamically optimizing and adjusting the driving path of the waste transport vehicle based on the Hamilton-Jacobi-Bellman equation combined with real-time traffic data; Multi-vehicle collaborative scheduling, optimizing multi-vehicle transportation scheduling based on differential game theory and using Nash equilibrium to solve the optimal driving path of each waste transport vehicle; Waste delivery and task completion, the waste transport vehicle drives to the target waste treatment station to unload waste and records the transportation data of this time to optimize subsequent transportation tasks.

Images