Intelligent container truck path planning control method, device, equipment and medium

Abstract

The invention discloses an intelligent container truck path planning control method and device, equipment and a medium, and relates to the field of path planning control, and the method comprises the steps: decomposing a path planning optimization problem into a plurality of independent time domain sub-blocks based on a time domain decomposition method, and obtaining a plurality of control periods of a prediction time domain; for each control period, based on the vehicle dynamics model, the state evolution equation and the operation state data, the current state and control input of the intelligent container truck are updated, and the updated state quantity and the updated control input are obtained; establishing a path planning sub-model of each time domain sub-block based on a Lagrange function, and updating the Lagrange dual variables to obtain updated dual variables; calculating a block moment according to the updated dual variable and the updated state quantity; determining whether a stop criterion is satisfied based on the block moment and the path performance index; and if yes, determining the current target trajectory as a final path planning result, and improving the calculation efficiency and the control precision.

Description

Technical Field

[0001] The present application relates to the field of path planning control, and in particular to a method, device, equipment and medium for intelligent container truck path planning control. Background Art

[0002] In the field of autonomous driving, path planning is particularly important for the operation of smart container trucks. However, the current smart container truck autonomous driving system has many deficiencies in path planning and control, especially when dealing with complex scenes. The traditional path planning method has low computational efficiency and control accuracy. Summary of the invention

[0003] The purpose of this application is to provide a smart container truck path planning control method, device, equipment and medium, which can improve the calculation efficiency and control accuracy of path planning.

[0004] To achieve the above objectives, this application provides the following solutions.

[0005] In a first aspect, the present application provides a method for intelligent container truck path planning and control, comprising the following steps: Obtain the operating status data of the smart container truck; Based on the time domain decomposition method, the path planning optimization problem is decomposed into multiple independent time domain sub-blocks to obtain several control cycles in the prediction time domain; one control cycle corresponds to one time domain sub-block; For each of the control cycles, based on the vehicle dynamics model, the state evolution equation and the operating state data, the current state and control input of the smart container truck are updated to obtain an updated state quantity and an updated control input; A path planning sub-model for each time-domain sub-block is established based on the Lagrangian function, and the Lagrangian dual variables are updated to obtain updated dual variables; block moments are calculated according to the updated dual variables and the updated state quantities; the path planning sub-model includes an objective function and constraints; the constraints include dynamic constraints and consistency constraints; the consistency constraints are determined by global consistency variables and are used to constrain the cooperative relationship between time-domain sub-blocks; the block moments are composed of state changes; Based on the block moment and the path performance index, it is judged whether the stopping criterion is met; if it is met, the current target trajectory is determined as the final path planning result; the path performance index includes a tracking performance index, a comfort performance index and a handling performance index; the stopping criterion is that the number of iterations reaches the maximum number of iterations, the tracking performance index reaches the maximum allowable tracking error, the comfort performance index reaches the maximum comfort performance deviation or the handling performance index reaches the maximum handling error.

[0006] In a second aspect, the present application provides an intelligent container truck path planning control device, comprising the following modules: The operation status data acquisition module is used to: obtain the operation status data of the smart container truck; A time domain decomposition module is used to: decompose the path planning optimization problem into multiple independent time domain sub-blocks based on the time domain decomposition method to obtain a number of control cycles in the prediction time domain; one control cycle corresponds to one time domain sub-block; The state evolution module is used to: for each control cycle, based on the vehicle dynamics model, the state evolution equation and the operating state data, update the current state and control input of the smart container truck to obtain an updated state quantity and an updated control input; The block update calculation module is used to: establish a path planning sub-model for each time domain sub-block based on the Lagrangian function, update the Lagrangian dual variables to obtain updated dual variables; calculate the block moment according to the updated dual variables and the updated state quantity; the path planning sub-model includes an objective function and constraints; the constraints include dynamic constraints and consistency constraints; the consistency constraints are determined by global consistency variables and are used to constrain the cooperative relationship between time domain sub-blocks; the block moment is composed of state changes; The stopping criterion convergence module is used to: determine whether the stopping criterion is met based on the block moment and the path performance index; if met, determine the current target trajectory as the final path planning result; the path performance index includes a tracking performance index, a comfort performance index and a handling performance index; the stopping criterion is that the number of iterations reaches the maximum number of iterations, the tracking performance index reaches the maximum allowable tracking error, the comfort performance index reaches the maximum comfort performance deviation or the handling performance index reaches the maximum handling error.

[0007] In a third aspect, the present application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-mentioned intelligent container truck path planning control method.

[0008] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-mentioned intelligent container truck path planning and control method.

[0009] According to the specific embodiments provided in this application, this application discloses the following technical effects: The present application provides a path planning and control method, device, equipment and medium for intelligent container trucks. The method decomposes the path planning optimization problem into multiple independent time domain sub-blocks to obtain several control cycles in the prediction time domain, calculates the path optimization results of each control cycle in parallel, synchronizes the states of each time domain sub-block through consistency variables, and finally achieves global optimization. The present application decomposes the complex global optimization problem into multiple sub-problems that can be processed in parallel through time domain decomposition and parallel computing technology, greatly reducing the calculation time and improving the calculation efficiency. By introducing a multi-dimensional performance index evaluation system, performance evaluation is performed in three aspects: tracking performance index, comfort performance index and handling performance index, comprehensively optimizing the path tracking performance and dynamic control capability of intelligent container trucks, and improving the control accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0011] Figure 1 This is an application environment diagram of an intelligent container truck path planning control method in an embodiment of the present application; Figure 2 A schematic diagram of a flow chart of a smart container truck path planning control method provided in one embodiment of the present application; Figure 3 A schematic diagram of a specific process of a smart container truck path planning control method provided in one embodiment of the present application; Figure 4 A schematic diagram of a constrained optimization quantization process based on time domain decomposition provided in an embodiment of the present application; Figure 5 A schematic diagram of a time domain decomposition process provided by an embodiment of the present application; Figure 6 A schematic diagram of functional modules of an intelligent container truck path planning control device provided in one embodiment of the present application; Figure 7 A schematic diagram of the structure of a computer device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0012] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0013] The existing path planning mainly has the following problems: First, the ability to process time-domain coupling constraints is insufficient. Existing technologies usually treat path planning as an overall optimization problem and fail to effectively decompose the coupling constraints in the time domain dimension. This processing method is rigid when dealing with multiple continuous control cycles and cannot flexibly adjust the state transition relationship between cycles, resulting in a significant reduction in computing efficiency. In addition, due to the lack of a complete time-domain decomposition mechanism, the potential of parallel computing cannot be fully utilized, which in turn restricts the system's ability to optimize in real time.

[0014] Secondly, the dynamic characteristics modeling is not comprehensive enough. The existing solutions fail to fully characterize the lateral and longitudinal dynamic characteristics of smart container trucks. Especially under complex working conditions, the prediction accuracy of the motion state of smart container trucks is obviously insufficient. In this case, the prediction error is magnified in practical applications, reducing the reliability of path planning. More importantly, there is currently no systematic method for quantitative evaluation of dynamic characteristics, which further limits the direction of technical optimization.

[0015] When dealing with global consistency constraints, existing path planning systems also have obvious technical shortcomings. Poor coordination of block calculations is one of the prominent problems. Existing technologies lack an efficient block decoupling mechanism, and it is difficult to achieve consistent coordination of states between different computing modules. Although some solutions attempt to introduce Lagrangian dual variables, the update strategy is not mature enough, resulting in low solution efficiency and even affecting the global optimization effect.

[0016] The lack of a performance indicator evaluation system further weakens the effectiveness of the system. Current solutions rarely involve H2 / Key performance indicators such as the norm and Lyapunov index cannot fully evaluate the quality of path planning solutions. The quantitative evaluation methods for power, comfort and maneuverability lack systematicity and fail to form a unified indicator framework, making performance evaluation difficult to standardize.

[0017] There is still much room for improvement in the existing technology for the stopping criteria of block optimization calculation. First, the evaluation index is single. Most of the existing stopping criteria focus on the evaluation of a single performance index, ignoring the comprehensive consideration of multi-dimensional factors such as vehicle spacing error and vehicle speed error. Especially under complex working conditions, this single-dimensional evaluation method is difficult to fully reflect the actual performance of the system. The current stopping criteria fail to effectively combine H2 / The reliability of the optimization results is questionable because the comprehensive performance evaluation is carried out based on the norm. Second, the dynamic adaptability is poor. The existing technology lacks the ability to adapt to the dynamic characteristics of the system, and the stopping criteria cannot automatically adjust the evaluation criteria according to the changes in different working conditions. The response to external disturbances is also relatively slow, making it difficult to meet the requirements of real-time and adaptability in complex operating environments.

[0018] Existing path planning technologies still face serious computational efficiency issues in practical applications, which further restricts their widespread application. First, the solution speed is insufficient: due to the failure to fully utilize the time domain decomposition technology, the overall computational burden is heavy, resulting in a slow convergence speed of path planning optimization. In addition, due to the imperfection of the parallel computing mechanism, the real-time control effect is often significantly affected. Secondly, the resource usage is high: the global optimization process requires a lot of computing resources, and the existing technology still has shortcomings in memory usage efficiency. This high resource usage characteristic not only makes it difficult to meet real-time control requirements, but also increases the hardware cost of system operation.

[0019] Existing path planning control technology has significant defects in time domain coupling constraint processing, global consistency constraint, stopping criterion design and computational efficiency. These problems seriously restrict the overall performance optimization of the intelligent container truck automatic driving system. To address the above problems, this application provides a path planning control method, device, equipment and medium for intelligent container trucks.

[0020] (1) Basic technical issues The core goal of this application is to solve the basic technical problems faced in the path planning of autonomous driving of intelligent container trucks, mainly focusing on the decoupling of time-domain coupling constraints and the optimization of block calculations. First of all, the decoupling problem of time-domain coupling constraints is a difficulty that needs to be solved urgently in current path planning technology. The existing technology is usually difficult to effectively deal with the time-domain coupling relationship between control cycles, and this application needs to achieve effective decoupling between different control cycles to ensure the independent optimization calculation of each control cycle. To this end, this application proposes a reasonable global consistency variable so that different time-domain blocks can work together efficiently and avoid the redundant calculation burden caused by time-domain coupling during the calculation process. At the same time, the decoupled subsystems must be able to ensure the convergence and stability of the solution, so as to avoid the uncertainty of the calculation results from having a negative impact on the system performance.

[0021] The optimization problem of block calculation is also one of the key points to be solved in this application. In order to improve the computational efficiency of path planning, this application proposes an efficient block update mechanism, which aims to optimize the solution process of each calculation block. In addition to ensuring the real-time and accuracy of data exchange between blocks, it is also necessary to balance the allocation of computing resources to ensure that system resources are not over-consumed while meeting performance requirements.

[0022] (2) Performance evaluation issues In view of the special needs of the intelligent container truck autonomous driving system in complex application scenarios, this application also needs to solve a series of performance evaluation problems. The core of these problems lies in how to comprehensively and accurately evaluate the performance of multiple dimensions such as power, comfort and controllability, and effectively combine them with the optimization goals. First of all, power performance evaluation is a key issue in path planning. How to use One of the important technical challenges of this application is to accurately evaluate the distance error and speed error. Based on this evaluation method, this application will also establish a reasonable tracking performance indicator system to ensure that the system's dynamic performance under different working conditions can be monitored and evaluated in real time, thereby ensuring the accuracy of path planning.

[0023] On this basis, the evaluation of comfort performance cannot be ignored. In order to achieve an accurate evaluation of driving comfort, this application uses the H2 norm for quantitative analysis, comprehensively considers the influence of control cost, external disturbance and other factors, and proposes a complete set of comfort quantitative evaluation standards. This standard not only reflects the stability of the control of smart trucks, but also effectively responds to the disturbance effects in various operating conditions.

[0024] In addition, the handling performance evaluation focuses on the evaluation of directional stability and trajectory error. By using the Lyapunov exponent, this application can effectively evaluate the directional stability of smart trucks and establish a complete trajectory error evaluation mechanism to ensure the continuous monitoring of handling performance in a real-time dynamic environment.

[0025] (3) Algorithm implementation issues To ensure the practical feasibility of the solution of this application, it is crucial to solve the algorithm implementation problem. First of all, computing efficiency optimization is the basis for improving system performance. This application proposes an efficient parallel computing strategy, which significantly improves the real-time response capability of the system by optimizing the iterative convergence speed and reducing resource consumption during the calculation process. In addition, the optimization of the iterative convergence speed is not only limited to improving the solution efficiency, but also involves how to reduce the occupancy of computing resources, thereby ensuring that the system can maintain stable operation under high load.

[0026] Secondly, the autonomous driving system must have strong robustness, especially in complex and dynamically changing environments. In order to cope with the changes and uncertainties in the external environment, this application proposes a robustness guarantee mechanism so that the algorithm can maintain stability under different working conditions and effectively resist the influence of external disturbances.

[0027] (4) Project implementation issues From the perspective of engineering implementation, this application also needs to solve several engineering and technical problems, especially in terms of real-time and reliability. First of all, real-time requirements are the key to the successful application of autonomous driving systems. How to ensure the real-time generation and execution of control instructions within the system has become an important issue that must be solved in this application. In response to this problem, this application proposes a set of efficient control mechanisms, which aims to reduce the impact of computing delays on control effects, and ensure the system's real-time response capabilities in various application scenarios by balancing control accuracy and real-time requirements.

[0028] In addition, reliability assurance also occupies an important position in engineering implementation. The long-term stability of the autonomous driving system is directly related to the safety and reliability of the system. Therefore, how to deal with potential constraints in system operation and establish an effective fault detection and handling mechanism under limited hardware resources has become an important task of this application. By designing redundant backup mechanisms and self-diagnosis functions, this application can ensure that the system remains stable during long-term operation and effectively respond to possible hardware or software failures.

[0029] (5) Comprehensive optimization problem This application also includes the solution of comprehensive optimization problems, especially in multi-objective coordination and enhanced adaptability. Autonomous driving path planning requires a balance between multiple performance indicators, especially power, comfort and maneuverability. How to achieve dynamic trade-offs between different control objectives to ensure the optimization of overall performance is an important technical challenge of this application. By introducing an adaptive adjustment mechanism, this application can flexibly adjust the weights of each performance indicator under different working conditions to ensure that the smart container truck can always maintain the best state in a dynamically changing environment.

[0030] Finally, enhanced adaptability is the key to improving system flexibility. The autonomous driving system needs to cope with various complex operating scenarios. Therefore, this application introduces an adaptive parameter adjustment mechanism so that the system can be adjusted in real time according to actual working conditions to achieve adaptation to different scenarios. Based on the above, this application ensures the wide applicability and high versatility of the algorithm, and can operate stably in a variety of environments.

[0031] By solving the above series of technical problems, this application aims to significantly improve the overall performance of the smart container truck autonomous driving system, thereby providing more reliable and efficient technical support for practical applications.

[0032] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0033] The intelligent container truck path planning control method provided in the embodiment of the present application can be applied to Figure 1In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be set up separately, integrated on the server 104, or placed on the cloud or other servers. The terminal 102 can send the running status data of the smart container truck to the server 104. After receiving the running status data, the server 104 decomposes the path planning optimization problem into multiple independent time domain sub-blocks based on the time domain decomposition method, and obtains several control cycles in the prediction time domain. For each control cycle, based on the vehicle dynamics model, the state evolution equation and the running status data, the current state and control input of the smart container truck are updated, and the path planning sub-model of each time domain sub-block is established based on the Lagrangian function, and the Lagrangian dual variables are updated to obtain the updated dual variables. According to the updated dual variables and the updated state quantity, the block moment is calculated, and the final path planning result is determined based on the block moment, the path performance index and the stopping criterion. The server 104 can feedback the final path planning result obtained for the smart container truck to the terminal 102. In addition, in some embodiments, the intelligent container truck path planning control method can also be implemented independently by the server 104 or the terminal 102. For example, the terminal 102 can directly perform path planning processing on the operating status data, or the server 104 can obtain the operating status data from the data storage system and perform path planning processing on the operating status data.

[0034] The terminal 102 may be, but is not limited to, various desktop computers and laptop computers. The server 104 may be implemented as an independent server or a server cluster consisting of multiple servers, or may be a cloud server.

[0035] In an exemplary embodiment, Figure 2 and Figure 3 As shown, a method for intelligent container truck path planning control is provided. The method is executed by a computer device, and can be executed by a computer device such as a terminal or a server alone, or by a terminal and a server together. In the embodiment of the present application, the method is applied to Figure 1 The server 104 in the example is used for explanation, and the steps include the following steps 201 to 205.

[0036] Step 201: Acquire the operation status data of the smart container truck. The operation status data includes the initial position, target path, real-time data, status information and environment information.

[0037] Step 202: Based on the time domain decomposition method, the path planning optimization problem is decomposed into multiple independent time domain sub-blocks to obtain a number of control cycles in the prediction time domain; one control cycle corresponds to one time domain sub-block.

[0038] Step 203: For each of the control cycles, based on the vehicle dynamics model, the state evolution equation and the operating state data, the current state and control input of the smart container truck are updated to obtain an updated state quantity and an updated control input.

[0039] Step 204: Establish a path planning sub-model for each time domain sub-block based on the Lagrangian function, update the Lagrangian dual variables to obtain updated dual variables; calculate the block moment according to the updated dual variables and the updated state quantity; the path planning sub-model includes an objective function and constraints; the constraints include dynamic constraints and consistency constraints; the consistency constraints are determined by global consistency variables and are used to constrain the collaborative relationship between time domain sub-blocks; the block moment is composed of state changes.

[0040] Step 205: judging whether the stopping criteria are met based on the block moment and the path performance index; if met, determining the current target trajectory as the final path planning result; the path performance index includes a tracking performance index, a comfort performance index and a handling performance index; the stopping criteria are that the number of iterations reaches the maximum number of iterations, the tracking performance index reaches the maximum allowable tracking error, the comfort performance index reaches the maximum comfort performance deviation or the handling performance index reaches the maximum handling error.

[0041] By implementing the above steps 201 to 205, the present application ensures the efficiency and accuracy of path planning through time domain decomposition, multi-dimensional performance index evaluation and block parallel computing technology. Specifically, the intelligent container truck path planning control method provided by the present application includes four core steps: time domain decomposition, global consistency variable construction, block update calculation, and performance evaluation and stop criterion judgment. These steps cooperate with each other to jointly achieve efficient optimization of path planning. In addition, the system architecture of the method is divided into three levels: input layer, processing layer and output layer. At the input layer, the system receives the operating status data of the intelligent container truck in real time, such as position, speed, acceleration, etc. At the processing layer, based on the time domain decomposition technology, the complex path planning problem is decomposed into multiple independent time domain sub-blocks, and the solution is optimized by parallel computing. At the output layer, path planning instructions that meet dynamic constraints and performance requirements are generated to ensure that the system can operate stably in complex scenarios.

[0042] like Figure 4 and Figure 5 As shown, the present application adopts a control cycle time domain decoupling method, and realizes a modular solution to the path planning problem by dividing the entire prediction time domain into multiple independent control cycles, which specifically includes the following steps 301 to 303.

[0043] Step 301: Time domain division: The entire prediction time domain is divided into N control cycles. The state of each control cycle is determined by the initial state quantity , input state quantity And the output state Representation.

[0044] Step 302: Dynamic constraint modeling: Based on the state transfer relationship obtained by time domain division in step (1) above, a dynamic constraint model is constructed to describe the state evolution of the system in each control cycle. The state evolution equation is shown as follows.

[0045] (1).

[0046] in, For the The output state quantity of a control cycle, For the The initial state quantity of a control cycle, For the The control input state quantity of a control cycle, is the state transfer matrix, is the control input matrix.

[0047] Step 303: Introduction of consistency variables: In order to ensure the global consistency of the block solution, this application introduces a global consistency variable , and establish the collaborative relationship between sub-blocks through consistency constraints. The consistency constraints include input consistency constraints and output consistency constraints ;in, For the Global consistency variables for each control cycle, For the The global consistency variables of each control cycle are composed of position state, velocity state and heading angle.

[0048] The final output of time domain decomposition is to divide the entire prediction time domain into multiple independent control cycles, establish a dynamic constraint model in each control cycle, and introduce global consistency variables to ensure the global consistency of block solution. Specifically, the output of this step includes: 1) Division of control cycles: Divide the prediction time domain into N control cycles, and the state quantities of each control cycle include initial state quantities, input state quantities, and output state quantities. 2) Dynamic constraint model: Construct a dynamic constraint model that describes the state evolution of the system in each control cycle. 3) Global consistency variables: Introduce global consistency variables, establish a collaborative relationship between sub-blocks through constraints, and realize time domain decoupling between control cycles.

[0049] The block decoupling calculation process is as follows: the state quantity includes the position information, speed and heading angle of the smart container truck; the control input includes the throttle, brake, acceleration and steering angular velocity. In each control cycle, the current state and control input are updated based on the vehicle dynamics model. Specifically, according to the current state of the smart container truck (including position, speed and acceleration, etc.) and control input (such as throttle, brake and steering, etc.), the smart container truck vehicle dynamics equation is used to calculate the state of the next control cycle. This process needs to meet the kinematic and dynamic constraints of the smart container truck, including the maximum acceleration constraint and maximum steering angular velocity constraint of the smart container truck. For example, if the current state is position and speed , the control input is acceleration , then the position state of the next control cycle and speed status It can be calculated by the following equation.

[0050] (2).

[0051] (3).

[0052] In the formula, For the The position state of a control cycle, For the The speed state of each control cycle, For the The position state of a control cycle, For the The speed state of each control cycle, is the time interval of the control cycle.

[0053] In path planning, the dual variable is used to determine the effective constraint, that is, when the dual variable is non-zero, the corresponding constraint is effective (active) or binding (binding). Use the Lagrangian dual variable to update the block state, optimize the objective function of each block, and ensure global coordination. The Lagrangian function contains the objective function and constraints, and the update of the dual variable needs to satisfy these constraints. For example, if the Lagrangian function is ,in, is a state variable, is the dual variable, then the update equation of the dual variable is as follows.

[0054] (4).

[0055] In the formula, is the updated dual variable, is the learning rate, is the gradient of the Lagrangian function with respect to the state variable.

[0056] The objective function is expressed as ,in, is the objective function, , and are the weights of the evaluation values ​​of the tracking performance index, the comfort performance index, and the handling performance index, respectively. The objective function aims at the minimum sum value, and the sum value is obtained by the index value of the tracking performance index. , Index values ​​of comfort performance index and the index value of the control performance index Calculated by weighted summation.

[0057] In each control cycle, the objective function and constraint conditions are iteratively solved to obtain the path planning result of each control cycle. The path planning result is evaluated by tracking performance indicators, comfort performance indicators and handling performance indicators. The final path planning result is determined based on the stopping criteria.

[0058] Initialization of primitive variables: At the beginning of each control cycle, the primitive variables are initialized according to the current state of the smart container truck. The primitive variables include the position, speed, acceleration and other state quantities of the smart container truck, as well as control inputs such as throttle, brake, steering, etc.

[0059] Construction of block vector: Based on the original variables and control inputs, a block vector representing the state of the current control cycle is constructed. A block vector is a collection of all variables related to path planning in a control cycle, including the state and control input of each control cycle.

[0060] The neighborhood characteristics are the environmental information of the smart container truck. Different neighborhood characteristics refer to the specific attributes of the smart container truck in different areas of its surrounding environment, such as obstacle distribution, road conditions, traffic rules, etc.

[0061] Calculation and optimization of residuals: In the path planning submodel, the residual between the actual path and the planned path is calculated. The residual is , and use the residual to update the Lagrange dual variables and block moments to optimize the path planning results. When the residual is not less than the set residual, continue the iterative optimization. When the residual is less than the set residual, stop the iteration and output the current path planning result.

[0062] The block moments include position change, vehicle spacing change, speed change, acceleration change, external disturbance change and trajectory error change.

[0063] Based on the updated state and the updated dual variables, the block moment is calculated and the input state of the next step is further adjusted. The calculation of the block moment involves changes in the system state, such as position change, speed change, etc. For example, when the block moment is a position change and speed changes When , it can be calculated by the following equation.

[0064] (5).

[0065] (6).

[0066] In order to achieve a comprehensive evaluation of path planning performance, this application proposes three types of path performance indicators: tracking performance, comfort performance and maneuverability performance.

[0067] 1) Tracking performance indicators: By measuring the distance error between vehicles and speed error ,Evaluate the tracking ability of the system to the target trajectory: In the time domain decomposition method, the state transition and dynamic constraint modeling of each control cycle directly affect the deviation between the actual trajectory of the smart truck vehicle and the target trajectory. Therefore, the tracking performance is closely related to the accuracy of the dynamic model and the division of the control cycle.

[0068] Performance evaluation logic: Pass The inter-vehicle distance error and vehicle speed error are quantified to evaluate the tracking performance of the system on the target trajectory in different control cycles.

[0069] (7).

[0070] in, To track the indicator value of the performance indicator; and are the weight coefficients of the vehicle spacing error and vehicle speed error respectively.

[0071] Norm calculation: By solving the integral of the tracking performance index, the inter-vehicle distance error and vehicle speed error can be obtained. norm to evaluate the tracking performance.

[0072] The error of vehicle spacing The calculation formula of the norm is shown below.

[0073] (8).

[0074] Speed ​​error The calculation formula of the norm is shown below.

[0075] (9).

[0076] The calculation formula of the evaluation value of the tracking performance indicator is shown in the following formula.

[0077] (10).

[0078] in, It's in time The vehicle spacing error, It's in time The speed error, The vehicle spacing error of norm; Speed ​​error of Norm.

[0079] 2) Comfort performance index: by analyzing external disturbances and control costs (acceleration change), evaluate the driving comfort of smart trucks: comfort performance is related to the control input (such as acceleration) and external disturbances (such as uneven road surface) in the vehicle dynamics model. The global consistency variables and block update calculations in the time domain decomposition method ensure the smooth transition of the state of the smart truck in different control cycles, which affects the comfort. Performance evaluation logic: Quantify the expected control cost and external disturbances through the H2 norm to evaluate the comfort performance of the system during driving.

[0080] In a specific example, when conducting a real vehicle test, the tracking error or speed error of the smart truck can be used as an external disturbance to evaluate For example, if the tracking error or speed error of a smart truck at a certain speed exceeds a preset threshold, it indicates that there is a significant external disturbance, such as an uneven road surface. The tracking error or speed error can be obtained through real-time measurements such as GPS data and speed sensors of smart trucks.

[0081] (11).

[0082] In the formula, is the index value of the comfort performance index; and are the weight coefficients of external disturbance and control cost respectively.

[0083] H2 norm calculation: By solving the integral of the comfort performance index, the H2 norm of the external disturbance and control cost can be obtained to evaluate the comfort performance.

[0084] External disturbance The calculation formula of the norm is shown below.

[0085] (12).

[0086] Control the cost The calculation formula of the norm is shown below.

[0087] (13).

[0088] The calculation formula of the comfort performance index is shown below.

[0089] (14).

[0090] in, It's in time The external disturbance It's in time The cost of control, For external disturbance of norm; To control the cost of Norm.

[0091] 3) Handling performance index: Based on the Lyapunov index, the rate of change of trajectory error is quantified to evaluate the handling stability of the system: The handling performance is closely related to the directional stability and trajectory error in the vehicle dynamics model. The time domain decomposition method can optimize the directional stability of smart trucks and reduce trajectory errors by introducing global consistency variables and block update calculations. Performance evaluation logic: The rate of change of trajectory error is quantified by the Lyapunov index to evaluate the handling performance of the system in a real-time dynamic environment.

[0092] Lyapunov exponent calculation: By solving the formula of the control performance index, the rate of change of the trajectory error can be obtained, thereby evaluating the control performance. The calculation formula of the Lyapunov exponent is shown in formula (15).

[0093] (15).

[0094] (16).

[0095] in, is the index value of the control performance index; is the rate of change of trajectory error; For current time t Trajectory error; is the initial error.

[0096] By calculating the Norm, The norm and Lyapunov exponent are compared with the preset performance index threshold to obtain the evaluation result of each path performance index.

[0097] Step 205 specifically includes: comparing the evaluation value of the path performance index with the preset performance index threshold to obtain the evaluation result of each path performance index; when the evaluation value of any path performance index is less than the preset performance index threshold, stopping the iterative optimization and outputting the final path planning result.

[0098] If the evaluation value of the tracking performance evaluation is less than the preset maximum allowable tracking error, the tracking performance is considered to meet the requirements; otherwise, the weight coefficient needs to be adjusted or the path planning needs to be optimized; if the evaluation value of the comfort performance evaluation is less than the preset maximum comfort performance deviation, the comfort performance is considered to meet the requirements; otherwise, the weight coefficient needs to be adjusted or the path planning needs to be optimized; if the evaluation value of the handling performance evaluation is less than the preset maximum handling error, the handling performance is considered to meet the requirements; otherwise, the weight coefficient needs to be adjusted or the path planning needs to be optimized. The maximum allowable tracking error, maximum comfort performance deviation and maximum handling error are obtained by experiments or user experience.

[0099] The present application may also use an LQR controller, mean absolute error (MAE) and root mean square error (RMSE), a multi-objective evolutionary algorithm performance evaluation index, a weighted sum method, a hierarchical analysis method, a fuzzy comprehensive evaluation method, a data envelopment analysis method, a non-dominated sorting genetic algorithm II (NSGA-II) or a multi-objective evolutionary algorithm based on decomposition (MOEA / D) for integral solution.

[0100] Comprehensive evaluation value It can reflect the overall performance of the system in terms of tracking performance, comfort performance and handling performance, and is used to balance the weights and influences between different performance indicators, thereby obtaining a comprehensive system performance evaluation. The calculation formula is shown below.

[0101] (17).

[0102] in, , and are the weights of the evaluation value of the tracking performance index, the evaluation value of the comfort performance, and the evaluation value of the handling performance, respectively.

[0103] Taking Port X as an example, the intelligent container truck path planning control method proposed in this application is specifically described below.

[0104] In the daily logistics operations of ports, straight sections usually connect cargo loading and unloading areas and container truck yards, and are an important part of logistics transportation. In this scenario, the smart container truck needs to travel on a 500-meter straight section with a target speed of 20 km / h. The task is to transport the cargo from the loading and unloading point to the yard, while dealing with potential dynamic interference such as forklifts and static obstacles such as containers. The port ground may be slippery or uneven due to environmental influences, and these external disturbances also require the system to make dynamic adjustments.

[0105] The process of obtaining the operating status data of the smart container truck is as follows: the initial position and target path of the smart container truck are provided by the port dispatching system, and the real-time data is obtained through the on-board sensors and V2X communication system. The status information is the current position coordinates of the smart container truck. , initial velocity , Target speed , Target location , maximum acceleration , and environmental information includes the location of static obstacles, the trajectory of dynamic targets, and ground conditions. In order to ensure the real-time and accuracy of the path planning results, a control cycle of 0.1 seconds and a prediction time domain of 3 seconds are used, that is, the state and control strategy of the smart truck are optimized within the next 30 control cycles.

[0106] The path planning uses time domain decomposition technology to divide the 3-second prediction time domain into 30 independent control cycles. In each control cycle, the state of the smart truck vehicle Describe the position information, speed and heading angle of the smart truck, and control variables Represents the acceleration and steering angular velocity of the smart truck. By introducing the global consistency variable , where the consistency variable for each control cycle is Include location status , speed status and heading angle , while ensuring that each block is solved independently, the consistency of the global optimization can also be maintained.

[0107] The state evolution of the intelligent truck is modeled based on the dynamic model. The state transition relationship is shown in formula (1), where the state transition matrix and the control input matrix , =0.1s is the time interval for each control cycle. After time domain decoupling, the constraints are divided into input consistency constraints and output consistency constraints. Dynamic modeling can describe the motion law of smart trucks on straight sections, while fully considering the impact of external disturbances on state changes.

[0108] In each time domain sub-block, the path planning algorithm performs the following steps. First, based on the vehicle dynamics model, the current state is updated and control input The calculation results must satisfy the kinematic and dynamic constraints of the intelligent truck. Then the block state is updated using the Lagrangian dual variables. The Lagrangian dual variables are combined with the dynamic constraints and consistency constraints to optimize the objective function of each block and ensure global coordination. Finally, based on the updated state and the updated dual variables, the block moment is calculated and the input state of the next step is further adjusted.

[0109] According to the specific requirements of the straight road section, the weights of the path performance indicators are set as follows: tracking performance has the highest priority, and the vehicle spacing error has the highest priority. Set to 0.7, speed error weight Set to 0.3. In order to balance comfort and handling performance, the perturbation weight and control cost weight Set to 0.4 and 0.6 respectively.

[0110] The optimization process adopts a block parallel calculation method, and each time domain sub-block solves the control strategy independently. The state of each block is synchronized through the consistency variable Z, and finally global optimization is achieved. In order to further improve the calculation efficiency, the system introduces an early stopping mechanism, and stops the iteration when the performance index reaches the preset threshold. The specific stopping criteria are: the maximum number of iterations is 100 times, and the maximum allowable tracking error is , maximum comfort performance deviation , maximum control error When any condition is met, the algorithm stops iterating and outputs the final path planning result. The performance indicators and test results are shown in Table 1.

[0111] Table 1 Performance indicators and test results

[0112] Through performance index analysis, tracking performance has been improved by 30%, comfort performance has been improved by about 25%, and handling stability has been improved by 35%. These results show that this application can not only achieve high-precision path tracking in complex environments, but also significantly optimize the driving experience, providing reliable technical support for port intelligent logistics.

[0113] Compared with the existing technology, this application shows significant advantages in computing efficiency, control accuracy, robustness and adaptability. Through time domain decomposition and constraint optimization methods, this application significantly improves the practical application performance of the path planning algorithm.

[0114] In terms of computing efficiency, this application effectively uses time-domain decomposition and parallel computing technology to decompose complex global optimization problems into multiple sub-problems that can be processed in parallel, greatly reducing the computing time. Test data shows that the computing time of this application is reduced by 40% to 50% compared with traditional methods, and the parallel computing efficiency is improved by 35% to 45%. In addition, by optimizing memory allocation and data flow processing, the system's memory resource usage is reduced by 30% to 40%, which is particularly outstanding in environments with limited hardware resources.

[0115] In terms of control accuracy, this application comprehensively optimizes the path tracking performance and dynamic control capabilities of smart container trucks by introducing a multi-dimensional performance index evaluation system. Experiments show that the smart container truck path planning control method proposed in this application reduces the path tracking error by 30% to 40%, improves the stability of smart container trucks by 25% to 35%, and improves the dynamic response speed by 20% to 30%. This improvement in accuracy and stability ensures that smart container trucks can smoothly and efficiently complete path planning and navigation tasks in complex environments.

[0116] In terms of robustness and adaptability, the application has significantly enhanced its resistance to external disturbances. Test results show that the application's anti-interference ability has been improved by 50%, and the control stability has been significantly improved in complex environments (such as dynamic obstacles, uneven or slippery roads). In addition, for different working conditions (such as curved paths, long straight paths, etc.), the application shows stronger adaptability and can quickly adjust algorithm parameters to meet the needs of different scenarios.

[0117] This application has achieved significant improvements in multiple key dimensions such as computational efficiency, control accuracy, and system robustness through time domain decomposition and global consistency variable construction methods. By introducing a multidimensional performance index evaluation system and an adaptive stopping criterion, this application not only solves the adaptability and efficiency bottlenecks of traditional methods in complex scenarios, but also proposes a set of scientific and systematic path planning quantitative evaluation methods. These technological breakthroughs provide a new technical solution for the path planning of autonomous driving of smart container trucks, which has important theoretical significance and application value. The technical solution of this application has shown great application value in many fields. This application can improve the path planning accuracy of autonomous driving of smart container trucks and reduce operating costs and energy consumption caused by planning errors.

[0118] In the field of industrial automation, the path planning technology and performance evaluation method of this application can provide theoretical support for the navigation and control of mobile robots in complex environments, and promote the development of autonomous driving technology in the industrial field.

[0119] The present application also provides an application scenario, which applies the above-mentioned intelligent container truck path planning control method. Specifically: The intelligent container truck path planning control method provided in this embodiment can be applied in an autonomous driving scenario. The autonomous driving scenario includes an information acquisition link, a path planning control link, and a driving link; the operating status data of the intelligent container truck enters the path planning control link from the information acquisition link, obtains the corresponding final path planning result, and enters the downstream driving link. The intelligent container truck path planning control method provided in this embodiment belongs to the path planning control link. Specifically, in the process of path planning control link for smart container trucks, the path planning optimization problem can be decomposed into multiple independent time domain sub-blocks based on the time domain decomposition method to obtain several control cycles in the prediction time domain. For each control cycle, the current state and control input of the smart container truck are updated based on the vehicle dynamics model, state evolution equation and operation status data. The path planning sub-model of each time domain sub-block is established based on the Lagrangian function, and the Lagrangian dual variables are updated to obtain updated dual variables. The block moments are calculated based on the updated dual variables and the updated state quantities, and the final path planning result is determined based on the block moments, path performance indicators and stopping criteria.

[0120] Based on the same inventive concept, the embodiment of the present application also provides an intelligent container truck path planning control device for implementing the intelligent container truck path planning control method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more intelligent container truck path planning control device embodiments provided below can refer to the limitations of the intelligent container truck path planning control method above, and will not be repeated here.

[0121] In an exemplary embodiment, Figure 6 As shown, an intelligent container truck path planning control device is provided, which includes the following modules.

[0122] The operation status data acquisition module T1 is used to obtain the operation status data of the smart container truck.

[0123] The time domain decomposition module T2 is used to: decompose the path planning optimization problem into multiple independent time domain sub-blocks based on the time domain decomposition method to obtain several control cycles in the prediction time domain; one control cycle corresponds to one time domain sub-block.

[0124] The state evolution module T3 is used to: for each control cycle, based on the vehicle dynamics model, the state evolution equation and the operating state data, update the current state and control input of the smart container truck to obtain an updated state quantity and an updated control input.

[0125] The block update calculation module T4 is used to: establish a path planning sub-model for each time domain sub-block based on the Lagrangian function, update the Lagrangian dual variables to obtain updated dual variables; calculate the block moment according to the updated dual variables and the updated state quantity; the path planning sub-model includes an objective function and constraints; the constraints include dynamic constraints and consistency constraints; the consistency constraints are determined by global consistency variables and are used to constrain the collaborative relationship between time domain sub-blocks; the block moment is composed of state changes.

[0126] The stopping criterion convergence module T5 is used to: determine whether the stopping criterion is met based on the block moment and the path performance index; if met, determine the current target trajectory as the final path planning result; the path performance index includes a tracking performance index, a comfort performance index and a handling performance index; the stopping criterion is that the number of iterations reaches the maximum number of iterations, the tracking performance index reaches the maximum allowable tracking error, the comfort performance index reaches the maximum comfort performance deviation or the handling performance index reaches the maximum handling error.

[0127] In an exemplary embodiment, a computer device is provided. The computer device may be a server or a terminal. The internal structure diagram thereof may be as follows: Figure 7 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. Among them, the processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store path planning control data. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for intelligent container truck path planning control is implemented.

[0128] Those skilled in the art will understand that Figure 7The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0129] In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above-mentioned method embodiments when executing the computer program.

[0130] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0131] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0132] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0133] The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. The non-relational database may include a distributed database based on blockchain, etc., but is not limited thereto. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but is not limited thereto.

[0134] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0135] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A smart container truck path planning and control method, characterized in that: The intelligent container truck path planning control method comprises: Obtain the operating status data of the smart container truck; Based on the time domain decomposition method, the path planning optimization problem is decomposed into multiple independent time domain sub-blocks to obtain several control cycles in the prediction time domain; one control cycle corresponds to one time domain sub-block; For each of the control cycles, based on the vehicle dynamics model, the state evolution equation and the operating state data, the current state and control input of the smart container truck are updated to obtain an updated state quantity and an updated control input; A path planning sub-model for each time-domain sub-block is established based on the Lagrangian function, and the Lagrangian dual variables are updated to obtain updated dual variables; block moments are calculated according to the updated dual variables and the updated state quantities; the path planning sub-model includes an objective function and constraints; the constraints include dynamic constraints and consistency constraints; the consistency constraints are determined by global consistency variables and are used to constrain the cooperative relationship between time-domain sub-blocks; the block moments are composed of state changes; Based on the block moment and the path performance index, it is judged whether the stopping criterion is met; if it is met, the current target trajectory is determined as the final path planning result; the path performance index includes a tracking performance index, a comfort performance index and a handling performance index; the stopping criterion is that the number of iterations reaches the maximum number of iterations, the tracking performance index reaches the maximum allowable tracking error, the comfort performance index reaches the maximum comfort performance deviation or the handling performance index reaches the maximum handling error.

2. The intelligent container truck path planning and control method according to claim 1 is characterized in that: The state evolution equation is expressed as follows: ； in, For the The output state quantity of a control cycle, For the The initial state quantity of a control cycle, For the The control input state quantity of a control cycle, is the state transfer matrix, is the control input matrix.

3. The intelligent container truck path planning and control method according to claim 1 is characterized in that: The consistency constraints include input consistency constraints and output consistency constraints ;in, For the The initial state quantity of a control cycle, For the Global consistency variables for each control cycle, For the The output state quantity of a control cycle, For the Global consistency variables for each control cycle.

4. The intelligent container truck path planning and control method according to claim 3 is characterized in that: The global consistency variables are composed of position state, velocity state and heading angle.

5. The intelligent container truck path planning and control method according to claim 1, characterized in that: The state variables include the position information, speed and heading angle of the smart container truck; the control inputs include acceleration and steering angular velocity.

6. The intelligent container truck path planning and control method according to claim 1, characterized in that: The block moments include position change, vehicle spacing change, speed change, acceleration change, external disturbance change and trajectory error change.

7. The intelligent container truck path planning and control method according to claim 1 is characterized in that: The calculation formula for the indicator value of the tracking performance indicator is as follows: ； in, To track the indicator value of the performance indicator; The vehicle spacing error of norm; Speed ​​error of norm; and are the weight coefficients of the vehicle spacing error and vehicle speed error respectively; The calculation formula of the index value of the comfort performance index is as follows: ； in, is the index value of the comfort performance index; For external disturbance of norm; To control the cost of norm; and are the weight coefficients of external disturbance and control cost respectively; The calculation formula of the index value of the maneuverability index is as follows: ； ； in, is the index value of the control performance index; is the rate of change of trajectory error; is the current trajectory error; is the initial error.

8. An intelligent container truck path planning control device, characterized in that: The intelligent container truck path planning control device comprises: The operation status data acquisition module is used to: obtain the operation status data of the smart container truck; A time domain decomposition module is used to: decompose the path planning optimization problem into multiple independent time domain sub-blocks based on the time domain decomposition method to obtain a number of control cycles in the prediction time domain; one control cycle corresponds to one time domain sub-block; The state evolution module is used to: for each control cycle, based on the vehicle dynamics model, the state evolution equation and the operating state data, update the current state and control input of the smart container truck to obtain an updated state quantity and an updated control input; The block update calculation module is used to: establish a path planning sub-model for each time domain sub-block based on the Lagrangian function, update the Lagrangian dual variables to obtain updated dual variables; calculate the block moment according to the updated dual variables and the updated state quantity; the path planning sub-model includes an objective function and constraints; the constraints include dynamic constraints and consistency constraints; the consistency constraints are determined by global consistency variables and are used to constrain the cooperative relationship between time domain sub-blocks; the block moment is composed of state changes; The stopping criterion convergence module is used to: determine whether the stopping criterion is met based on the block moment and the path performance index; if met, determine the current target trajectory as the final path planning result; the path performance index includes a tracking performance index, a comfort performance index and a handling performance index; the stopping criterion is that the number of iterations reaches the maximum number of iterations, the tracking performance index reaches the maximum allowable tracking error, the comfort performance index reaches the maximum comfort performance deviation or the handling performance index reaches the maximum handling error.

9. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the intelligent container truck path planning control method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the intelligent container truck path planning control method described in any one of claims 1 to 7 is implemented.

Images