Intersection cooperative passing method and system based on port logistics L4 unmanned container trucks

Abstract

The invention relates to an intersection cooperative passing method and system based on a port logistics L4 unmanned container truck, and the method comprises the steps: U1, enabling the unmanned container truck to execute an operation task at a port, collecting the driving track data information of the historical operation of a vehicle, obtaining the position data information of the vehicle in real time based on a vehicle-mounted positioning system, and obtaining the position data information of the vehicle; acquiring traffic condition data information of the intersection of the current road in real time based on a road side unit (RSU); and U2, based on the driving track data information of the historical operation of the vehicle and the position data information of the vehicle, adopting an improved crown porcupine optimization-convolution-bidirectional long and short term memory network-attention mechanism data regression prediction algorithm to predict a vehicle driving track point sequence. According to the method, the overall automation level can be provided, and adaptive adjustment and planning can be carried out, so that the fine control requirement of business is met, the overall operation efficiency is improved, the operation income is increased, and the development and operation human input cost is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned container trucks, and in particular to a method and system for coordinated intersection passage based on port logistics L4 unmanned container trucks. Background Art

[0002] The existing port truck intersection coordination method abstracts an area extending outward from the intersection into a control zone. When multiple vehicles arrive at the control zone, it determines whether there will be a conflict when the vehicles pass through the intersection by judging whether the future driving trajectories reported by the autonomous driving will intersect. If multiple vehicles conflict, the first vehicle to pass through the control zone is allowed to pass first, and the subsequent vehicles queue up and wait. Otherwise, multiple vehicles pass through the intersection at the same time.

[0003] This method can solve the conflict problem when multiple vehicles pass through the intersection, but it will affect the traffic efficiency. For example, when the vehicle that first enters the control area but has not yet entered the intersection conflicts with multiple subsequent vehicles that do not conflict with each other, the subsequent vehicles will be asked to stop and wait due to the first-come-first-go principle. At this time, the traffic efficiency of a single intersection is not optimal, because the impact of allowing multiple subsequent vehicles that do not conflict with each other to pass first will be greatly reduced. Secondly, it is impossible to customize the traffic priority according to the business. Since the existing conflict judgment is based on the trajectory points to determine whether there is a conflict, and the direction of the vehicle is unknown, if you want to achieve refined control such as giving priority to vehicles in the north-south or east-west directions or vehicles turning a large curve, the existing method cannot meet the requirements. Summary of the invention

[0004] In view of the above problems, the present invention provides a method and system for coordinated intersection passage based on port logistics L4 unmanned container trucks, which can not only provide an overall level of automation, but also perform adaptive adjustment and planning, thereby meeting the needs of refined business control, improving overall operational efficiency, increasing operating income, and reducing development and operation manpower investment costs.

[0005] In order to achieve the above-mentioned purpose and other related purposes, the technical solution provided by the present invention is as follows:

[0006] A method for coordinated crossing of an intersection based on port logistics L4 unmanned container trucks, the method comprising:

[0007] U1. Unmanned container trucks perform operations at ports, collect historical operation trajectory data of vehicles, obtain vehicle location data in real time based on the vehicle positioning system, and obtain traffic status data at intersections of the current road in real time based on the roadside unit RSU;

[0008] U2. Based on the driving trajectory data information of the historical operation of the vehicle and the location data information of the vehicle, the vehicle driving trajectory point sequence is predicted using an improved crown porcupine optimization-convolution-bidirectional long short-term memory network-attention mechanism data regression prediction algorithm to obtain the predicted vehicle driving trajectory point sequence data information;

[0009] U3. Based on the predicted vehicle driving trajectory point sequence data information and the traffic condition data information of the intersection of the current road, the vehicle driving decision data information is predicted using a prediction algorithm based on an improved gray wolf optimized BP neural network with multiple inputs and multiple outputs to obtain the predicted vehicle driving decision data information;

[0010] U4. Based on the predicted vehicle driving decision data information, a feasibility function F of the coordinated passage of vehicles at intersections is established, and the feasibility value of the coordinated passage of vehicles at intersections is calculated to obtain the feasibility value data information of the coordinated passage of vehicles at intersections.

[0011] Furthermore, in step U2, the improved crown porcupine optimization-convolution-bidirectional long short-term memory network-attention mechanism data regression prediction algorithm predicts the vehicle driving trajectory point sequence including:

[0012] U21. Input the driving trajectory data information of the vehicle's historical operation into the convolutional neural network for local feature extraction to obtain the driving trajectory point feature matrix data information of the vehicle's historical operation, input the vehicle's position data information into the convolutional neural network for local feature extraction to obtain the vehicle's position point feature matrix data information;

[0013] U22. Input the feature matrix data information of the vehicle's position points and the feature matrix data information of the vehicle's historical operation driving trajectory points into a bidirectional long short-term memory network to characterize the correlation between the vehicle's position points and the vehicle's historical operation driving trajectory points, and obtain the correlation data information between the vehicle's position points and the historical operation driving trajectory points;

[0014] U23. Based on the correlation data information between the vehicle position point and the driving trajectory point of the historical operation, establish the driving trajectory point sequence prediction function Q of the vehicle after the crown porcupine optimization,

[0015]

[0016] Among them, x is the correlation data information between the vehicle position point and the driving trajectory point of the historical operation, α1, α2 and α3 are the weight factors optimized by the crown porcupine, and the vehicle driving trajectory point sequence is predicted to obtain the predicted vehicle driving trajectory point sequence data information.

[0017] Furthermore, the constraints of the optimized weight factors α1, α2 and α3 of the crested porcupine are:

[0018]

[0019] The weight factor after the crown porcupine optimization is obtained by optimizing the initial weight factor using an improved crown porcupine optimization algorithm to obtain the weight factor after the crown porcupine optimization.

[0020] Furthermore, the optimization of the initial weight factor using the improved crested porcupine optimization algorithm includes:

[0021] U231. Based on the initial weight factor, the crested porcupine population is initialized to obtain the initialized crested porcupine population data information;

[0022] U232. Based on the initialized crested porcupine population data information, establish the fitness function W of the population individuals,

[0023]

[0024] Among them, y is the initialized crested porcupine population data information, β1 and β2 are the fitness difference factors of the population individuals, and the fitness values ​​of the population individuals are calculated to obtain the fitness value data information of the population individuals;

[0025] U233. Based on the fitness value data information of the population individuals, establish the target optimization function R of the population individuals,

[0026]

[0027] Among them, z is the fitness value data information of the population individuals, δ1, δ2 and δ3 are the optimization factors of the population individuals, and the initial weight factors are optimized to obtain the optimized weight factors of the crown porcupine.

[0028] Furthermore, in step U233, the constraint function f of the optimization factors δ1, δ2 and δ3 of the population individuals is,

[0029]

[0030] The value range of the constraint function f is (2, 3).

[0031] Furthermore, in step U3, the prediction of the vehicle's driving decision data information using a prediction algorithm based on an improved grey wolf optimized BP neural network with multiple inputs and multiple outputs includes:

[0032] U31. Based on the predicted vehicle driving trajectory point sequence data information and the traffic condition data information of the current road intersection, the BP neural network is initialized to obtain the network parameter data information of the initialized BP neural network;

[0033] U32. Based on the network parameter data information of the initialized BP neural network, the gray wolf population is initialized to obtain the initialized gray wolf population data information, and the fitness function P of the gray wolf population individuals is constructed.

[0034]

[0035] Among them, r is the initialized gray wolf population information, γ1 and γ2 are the fitness determining factors of the gray wolf population individuals, and the individual fitness values ​​of the gray wolf population are calculated to obtain the fitness value data information of the gray wolf population individuals;

[0036] U33. Based on the fitness value data information of the gray wolf population individuals, establish the objective function S,

[0037]

[0038] Among them, a is the fitness value data information of the gray wolf population individuals, η1, η2 and η3 are the target optimization constant parameters of the gray wolf population, and the network parameters of the BP neural network are optimized to obtain the optimized BP neural network model of the gray wolf;

[0039] U34. Based on the BP neural network model optimized by the Gray Wolf, the predicted vehicle driving trajectory point sequence data information and the traffic condition data information of the intersection of the current road are input to predict the vehicle's driving decision data information to obtain the predicted vehicle's driving decision data information.

[0040] Furthermore, the constraints of the target optimization constant parameters η1, η2 and η3 of the gray wolf population are:

[0041]

[0042] Furthermore, the feasibility function F of the vehicle intersection cooperative passage is:

[0043]

[0044] Among them, c is the predicted driving decision data information of the vehicle, λ1, λ2 and λ3 are the adaptive adjustment factors of the intersection cooperative passing;

[0045] The method further comprises:

[0046] U5. Based on the feasibility value data information of the vehicle's coordinated passage through the intersection, a preset threshold is set. If the feasibility value data of the vehicle's coordinated passage through the intersection is less than the preset threshold, the unmanned container truck can safely pass through the intersection. If the feasibility value data of the vehicle's coordinated passage through the intersection is greater than the preset threshold, the unmanned container truck cannot pass through the intersection.

[0047] Furthermore, the adaptive adjustment factors λ1, λ2 and λ3 of the intersection cooperative passage are:

[0048]

[0049] Among them, c is the predicted driving decision data information of the vehicle.

[0050] In order to achieve the above-mentioned objectives and other related objectives, the present invention also provides a coordinated passage system for intersections based on port logistics L4 unmanned container trucks, including a computer device, which is programmed or configured to execute any one of the steps of the coordinated passage method for intersections based on port logistics L4 unmanned container trucks.

[0051] The present invention has the following positive effects:

[0052] 1. The present invention predicts the vehicle driving trajectory point sequence by adopting an improved crown porcupine optimization-convolution-bidirectional long short-term memory network-attention mechanism data regression prediction algorithm, which can not only accurately predict the vehicle's driving trajectory points, but also the entire process does not require human participation, thereby reducing a large amount of human resource consumption and improving the operating efficiency of unmanned container trucks in ports.

[0053] 2. The present invention predicts the vehicle's driving decision data information by adopting a prediction algorithm based on the improved gray wolf optimized BP neural network with multiple inputs and multiple outputs, and combines the feasibility function F of the coordinated passage of vehicles at intersections to calculate the feasibility value of the coordinated passage of vehicles at intersections. It can not only provide an overall automation level, but also perform adaptive adjustment and planning, thereby meeting the needs of refined business control, improving overall operational efficiency, increasing operating income, and reducing development and operation manpower investment costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 It is a schematic diagram of the method flow of the present invention;

[0055] Figure 2 It is a flow chart of the improved crested porcupine optimization-convolution-bidirectional long short-term memory network-attention mechanism data regression prediction algorithm of the present invention;

[0056] Figure 3 It is a flow chart of the prediction algorithm based on improved gray wolf optimized BP neural network with multiple input and multiple output of the present invention;

[0057] Figure 4 It is a schematic diagram of the flow chart of the improved crested porcupine optimization algorithm of the present invention. DETAILED DESCRIPTION

[0058] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0059] Example 1: Figure 1 As shown, a method for coordinated crossing of an intersection based on port logistics L4 unmanned container trucks, the method comprising:

[0060] U1. Unmanned container trucks perform operations at the port, collect historical operation trajectory data of the vehicle, obtain the vehicle's location data in real time based on the vehicle positioning system, and obtain the traffic status data of the current road intersection in real time based on the roadside unit RSU;

[0061] U2. Based on the driving trajectory data information of the historical operation of the vehicle and the location data information of the vehicle, the vehicle driving trajectory point sequence is predicted using an improved crown porcupine optimization-convolution-bidirectional long short-term memory network-attention mechanism data regression prediction algorithm to obtain the predicted vehicle driving trajectory point sequence data information;

[0062] U3. Based on the predicted vehicle driving trajectory point sequence data information and the traffic condition data information of the intersection of the current road, the vehicle driving decision data information is predicted using a prediction algorithm based on an improved gray wolf optimized BP neural network with multiple inputs and multiple outputs to obtain the predicted vehicle driving decision data information;

[0063] U4. Based on the predicted vehicle driving decision data information, a feasibility function F of the coordinated passage of vehicles at intersections is established, and the feasibility value of the coordinated passage of vehicles at intersections is calculated to obtain the feasibility value data information of the coordinated passage of vehicles at intersections.

[0064] In this embodiment, if Figure 2 As shown, in step U2, the improved crown porcupine optimization-convolution-bidirectional long short-term memory network-attention mechanism data regression prediction algorithm predicts the vehicle driving trajectory point sequence including:

[0065] U21. Input the driving trajectory data information of the vehicle's historical operation into the convolutional neural network for local feature extraction to obtain the driving trajectory point feature matrix data information of the vehicle's historical operation, input the vehicle's position data information into the convolutional neural network for local feature extraction to obtain the vehicle's position point feature matrix data information;

[0066] U22. Input the feature matrix data information of the vehicle's position points and the feature matrix data information of the vehicle's historical operation driving trajectory points into a bidirectional long short-term memory network to characterize the correlation between the vehicle's position points and the vehicle's historical operation driving trajectory points, and obtain the correlation data information between the vehicle's position points and the historical operation driving trajectory points;

[0067] U23. Based on the correlation data information between the vehicle position point and the driving trajectory point of the historical operation, establish the driving trajectory point sequence prediction function Q of the vehicle after the crown porcupine optimization,

[0068]

[0069] Among them, x is the correlation data information between the vehicle position point and the driving trajectory point of the historical operation, α1, α2 and α3 are the weight factors optimized by the crown porcupine, and the vehicle driving trajectory point sequence is predicted to obtain the predicted vehicle driving trajectory point sequence data information.

[0070] In this embodiment, the constraints of the optimized weight factors α1, α2 and α3 of the crested porcupine are:

[0071]

[0072] The weight factor after the crown porcupine optimization is obtained by optimizing the initial weight factor using an improved crown porcupine optimization algorithm to obtain the weight factor after the crown porcupine optimization.

[0073] In this embodiment, if Figure 4 As shown, the optimization of the initial weight factor using the improved crown porcupine optimization algorithm includes:

[0074] U231. Based on the initial weight factor, the crested porcupine population is initialized to obtain the initialized crested porcupine population data information;

[0075] U232. Based on the initialized crested porcupine population data information, establish the fitness function W of the population individuals,

[0076]

[0077] Among them, y is the initialized crested porcupine population data information, β1 and β2 are the fitness difference factors of the population individuals, and the fitness values ​​of the population individuals are calculated to obtain the fitness value data information of the population individuals;

[0078] U233. Based on the fitness value data information of the population individuals, establish the target optimization function R of the population individuals,

[0079]

[0080] Among them, z is the fitness value data information of the population individuals, δ1, δ2 and δ3 are the optimization factors of the population individuals, and the initial weight factors are optimized to obtain the optimized weight factors of the crown porcupine.

[0081] In this embodiment, in step U233, the constraint function f of the optimization factors δ1, δ2 and δ3 of the population individuals is,

[0082]

[0083] The value range of the constraint function f is (2, 3).

[0084] In this embodiment, intersection collaborative control is divided into four layers, namely data access domain, data identification domain, rule engine decision domain and execution domain. The terminal collection program on the unmanned container truck and port machinery equipment transmits the signal to the data collection gateway through the 5G / dedicated line network. The data collection gateway persists the data to the message queue. The data identification domain extracts the vehicle latitude and longitude, the global path of the current task, the high-precision map intersection & path correspondence, etc. by consuming the message queue data, identifies the spatial relationship between the vehicle and the intersection, the sub-intersection information of the vehicle entering and exiting the fence, the vehicle driving direction, etc., and compares the vehicles to be driven into the same intersection in pairs and divides them into conflict & non-conflict groups. The rule engine decision domain decides the passage priority of vehicles in different non-conflict groups according to the built-in and business-defined priority rules. The execution domain receives the decision result of the rule engine domain and converts it into the corresponding decision instruction and sends it to the vehicle end for execution, completing the vehicle start and stop control and ending a round of decision making.

[0085] Example 2: Based on the coordinated intersection passing method of port logistics L4 unmanned container trucks in Example 1, the present invention is further illustrated and described below.

[0086] like Figure 1 As shown, a method for coordinated crossing of an intersection based on port logistics L4 unmanned container trucks, the method comprising:

[0087] U1. Unmanned container trucks perform operations at ports, collect historical operation trajectory data of vehicles, obtain vehicle location data in real time based on the vehicle positioning system, and obtain traffic status data at intersections of the current road in real time based on the roadside unit RSU;

[0088] U2. Based on the driving trajectory data information of the historical operation of the vehicle and the location data information of the vehicle, the vehicle driving trajectory point sequence is predicted using an improved crown porcupine optimization-convolution-bidirectional long short-term memory network-attention mechanism data regression prediction algorithm to obtain the predicted vehicle driving trajectory point sequence data information;

[0089] U3. Based on the predicted vehicle driving trajectory point sequence data information and the traffic condition data information of the intersection of the current road, the vehicle driving decision data information is predicted using a prediction algorithm based on an improved gray wolf optimized BP neural network with multiple inputs and multiple outputs to obtain the predicted vehicle driving decision data information;

[0090] U4. Based on the predicted vehicle driving decision data information, a feasibility function F of the coordinated passage of vehicles at intersections is established, and the feasibility value of the coordinated passage of vehicles at intersections is calculated to obtain the feasibility value data information of the coordinated passage of vehicles at intersections.

[0091] In this embodiment, if Figure 3 As shown, in step U3, the prediction of the vehicle's driving decision data information using the improved grey wolf optimized BP neural network multi-input multi-output prediction algorithm includes:

[0092] U31. Based on the predicted vehicle driving trajectory point sequence data information and the traffic condition data information of the current road intersection, the BP neural network is initialized to obtain the network parameter data information of the initialized BP neural network;

[0093] U32. Based on the network parameter data information of the initialized BP neural network, the gray wolf population is initialized to obtain the initialized gray wolf population data information, and the fitness function P of the gray wolf population individuals is constructed.

[0094]

[0095] Among them, r is the initialized gray wolf population information, γ1 and γ2 are the fitness determining factors of the gray wolf population individuals, and the individual fitness values ​​of the gray wolf population are calculated to obtain the fitness value data information of the gray wolf population individuals;

[0096] U33. Based on the fitness value data information of the gray wolf population individuals, establish the objective function S,

[0097]

[0098] Among them, a is the fitness value data information of the gray wolf population individuals, η1, η2 and η3 are the target optimization constant parameters of the gray wolf population, and the network parameters of the BP neural network are optimized to obtain the optimized BP neural network model of the gray wolf;

[0099] U34. Based on the BP neural network model optimized by the Gray Wolf, the predicted vehicle driving trajectory point sequence data information and the traffic condition data information of the intersection of the current road are input to predict the vehicle's driving decision data information to obtain the predicted vehicle's driving decision data information.

[0100] In this embodiment, the constraint conditions of the target optimization constant parameters η1, η2 and η3 of the gray wolf population are:

[0101]

[0102] In this embodiment, the feasibility function F of the vehicle intersection cooperative passage is:

[0103]

[0104] Among them, c is the predicted driving decision data information of the vehicle, λ1, λ2 and λ3 are the adaptive adjustment factors of the intersection cooperative passing;

[0105] The method further comprises:

[0106] U5. Based on the feasibility value data information of the vehicle's coordinated passage through the intersection, a preset threshold is set. If the feasibility value data of the vehicle's coordinated passage through the intersection is less than the preset threshold, the unmanned container truck can safely pass through the intersection. If the feasibility value data of the vehicle's coordinated passage through the intersection is greater than the preset threshold, the unmanned container truck cannot pass through the intersection.

[0107] In this embodiment, the adaptive adjustment factors λ1, λ2 and λ3 of the intersection cooperative passage are:

[0108]

[0109] Among them, c is the predicted driving decision data information of the vehicle.

[0110] In this embodiment, the present invention provides a coordinated passage system for intersections based on port logistics L4 unmanned container trucks, including a computer device that is programmed or configured to execute any one of the steps of the coordinated passage method for intersections based on port logistics L4 unmanned container trucks.

[0111] The present invention also provides a computer-readable storage medium, which stores a computer program programmed or configured to execute any one of the methods for coordinated intersection passage based on port logistics L4 unmanned container trucks.

[0112] Any reference to memory, storage, database or other media used in the embodiments provided in the present application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM) and memory bus dynamic RAM (RDRAM).

[0113] In summary, the present invention can not only provide an overall automation level, but also perform adaptive adjustment and planning, so as to meet the needs of refined business control, improve overall operational efficiency, increase operating income, and reduce development and operation manpower investment costs.

[0114] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A method for coordinated crossing of L4 unmanned container trucks in port logistics, characterized in that: The method comprises: U1. Unmanned container trucks perform operations at the port, collect historical operation trajectory data of the vehicle, obtain the vehicle's location data in real time based on the vehicle positioning system, and obtain the traffic status data of the current road intersection in real time based on the roadside unit RSU; U2. Based on the driving trajectory data information of the historical operation of the vehicle and the location data information of the vehicle, the vehicle driving trajectory point sequence is predicted using an improved crown porcupine optimization-convolution-bidirectional long short-term memory network-attention mechanism data regression prediction algorithm to obtain the predicted vehicle driving trajectory point sequence data information; U3. Based on the predicted vehicle driving trajectory point sequence data information and the traffic condition data information of the intersection of the current road, the vehicle driving decision data information is predicted using a prediction algorithm based on an improved gray wolf optimized BP neural network with multiple inputs and multiple outputs to obtain the predicted vehicle driving decision data information; U4. Based on the predicted vehicle driving decision data information, a feasibility function F of the coordinated passage of vehicles at intersections is established, and the feasibility value of the coordinated passage of vehicles at intersections is calculated to obtain the feasibility value data information of the coordinated passage of vehicles at intersections.

2. The method for coordinated crossing of an intersection based on port logistics L4 unmanned container trucks according to claim 1 is characterized in that: In step U2, the improved crown porcupine optimization-convolution-bidirectional long short-term memory network-attention mechanism data regression prediction algorithm predicts the vehicle driving trajectory point sequence including: U21. Input the driving trajectory data information of the vehicle's historical operation into the convolutional neural network for local feature extraction to obtain the driving trajectory point feature matrix data information of the vehicle's historical operation, input the vehicle's position data information into the convolutional neural network for local feature extraction to obtain the vehicle's position point feature matrix data information; U22. Input the feature matrix data information of the vehicle's position points and the feature matrix data information of the vehicle's historical operation driving trajectory points into a bidirectional long short-term memory network to characterize the correlation between the vehicle's position points and the vehicle's historical operation driving trajectory points, and obtain the correlation data information between the vehicle's position points and the historical operation driving trajectory points; U23. Based on the correlation data information between the vehicle position point and the driving trajectory point of the historical operation, establish the driving trajectory point sequence prediction function Q of the vehicle after the crown porcupine optimization, Among them, x is the correlation data information between the vehicle position point and the driving trajectory point of the historical operation, α1, α2 and α3 are the weight factors optimized by the crown porcupine, and the vehicle driving trajectory point sequence is predicted to obtain the predicted vehicle driving trajectory point sequence data information.

3. The method for coordinated crossing of an intersection based on port logistics L4 unmanned container trucks according to claim 2 is characterized in that: The constraints of the optimized weight factors α1, α2 and α3 of the crested porcupine are: The weight factor after the crown porcupine optimization is obtained by optimizing the initial weight factor using an improved crown porcupine optimization algorithm to obtain the weight factor after the crown porcupine optimization.

4. The method for coordinated crossing of an intersection based on port logistics L4 unmanned container trucks according to claim 3 is characterized in that: The method of optimizing the initial weight factor by using the improved crested porcupine optimization algorithm includes: U231. Based on the initial weight factor, the crested porcupine population is initialized to obtain the initialized crested porcupine population data information; U232. Based on the initialized crested porcupine population data information, establish the fitness function W of the population individuals, Among them, y is the initialized crested porcupine population data information, β1 and β2 are the fitness difference factors of the population individuals, and the fitness values ​​of the population individuals are calculated to obtain the fitness value data information of the population individuals; U233. Based on the fitness value data information of the population individuals, establish the target optimization function R of the population individuals, Among them, z is the fitness value data information of the population individuals, δ1, δ2 and δ3 are the optimization factors of the population individuals, and the initial weight factors are optimized to obtain the optimized weight factors of the crown porcupine.

5. The method for coordinated crossing of an intersection based on port logistics L4 unmanned container trucks according to claim 4 is characterized in that: In step U233, the constraint function f of the optimization factors δ1, δ2 and δ3 of the population individuals is, The value range of the constraint function f is (2, 3).

6. The method for coordinated crossing of an intersection based on port logistics L4 unmanned container trucks according to claim 1 is characterized in that: In step U3, the prediction of the vehicle's driving decision data information using a prediction algorithm based on an improved grey wolf optimized BP neural network with multiple inputs and multiple outputs includes: U31. Based on the predicted vehicle driving trajectory point sequence data information and the traffic condition data information of the current road intersection, the BP neural network is initialized to obtain the network parameter data information of the initialized BP neural network; U32. Based on the network parameter data information of the initialized BP neural network, the gray wolf population is initialized to obtain the initialized gray wolf population data information, and the fitness function P of the gray wolf population individuals is constructed. Among them, r is the initialized gray wolf population information, γ1 and γ2 are the fitness determining factors of the gray wolf population individuals, and the individual fitness values ​​of the gray wolf population are calculated to obtain the fitness value data information of the gray wolf population individuals; U33. Based on the fitness value data information of the gray wolf population individuals, establish the objective function S, Among them, a is the fitness value data information of the gray wolf population individuals, η1, η2 and η3 are the target optimization constant parameters of the gray wolf population, and the network parameters of the BP neural network are optimized to obtain the optimized BP neural network model of the gray wolf; U34. Based on the BP neural network model optimized by the Gray Wolf, the predicted vehicle driving trajectory point sequence data information and the traffic condition data information of the intersection of the current road are input to predict the vehicle's driving decision data information to obtain the predicted vehicle's driving decision data information.

7. The method for coordinated crossing of an intersection based on port logistics L4 unmanned container trucks according to claim 6 is characterized by: The constraints of the target optimization constant parameters η1, η2 and η3 of the gray wolf population are:

8. The method for coordinated crossing of an intersection based on port logistics L4 unmanned container trucks according to claim 1 is characterized by: The feasibility function F of the vehicle intersection cooperative passage is: Among them, c is the predicted driving decision data information of the vehicle, λ1, λ2 and λ3 are the adaptive adjustment factors of the intersection cooperative passing; The method further comprises: U5. Based on the feasibility value data information of the vehicle's coordinated passage through the intersection, a preset threshold is set. If the feasibility value data of the vehicle's coordinated passage through the intersection is less than the preset threshold, the unmanned container truck can safely pass through the intersection. If the feasibility value data of the vehicle's coordinated passage through the intersection is greater than the preset threshold, the unmanned container truck cannot pass through the intersection.

9. The method for coordinated crossing of an intersection based on port logistics L4 unmanned container trucks according to claim 8 is characterized by: The adaptive adjustment factors λ1, λ2 and λ3 of the intersection cooperative passage are: Among them, c is the predicted driving decision data information of the vehicle.

10. A coordinated intersection passing system based on port logistics L4 unmanned container trucks, including computer equipment, characterized in that: The computer device is programmed or configured to execute the steps of the intersection collaborative passage method based on port logistics L4 unmanned container trucks as described in any one of claims 1 to 9.

Images