Automatic driving vehicle group pick-up optimization system based on group behavior prediction

Abstract

The invention discloses an automatic driving vehicle group pick-up and drop-off optimization system based on group behavior prediction. The system comprises a passenger group data acquisition module, a group behavior prediction and mode analysis module, a multi-mode large model optimization module, a group pick-up and drop-off point priority ranking and dynamic adjustment module and a real-time group interaction and feedback module. Through real-time analysis of behaviors, demands, positions and future movement tracks of a plurality of passengers and in combination with a group behavior prediction technology, the automatic driving vehicle can dynamically adjust pick-up points and driving routes, the pick-up efficiency and passenger experience are maximized, and the passenger experience is improved. The system is especially suitable for sharing travel, multi-passenger car sharing and dynamically changing multi-person pick-up and drop-off scenes.

Description

Technical Field

[0001] The present invention relates to the field of autonomous driving technology, and in particular to an optimization system for group pick-up and drop-off of autonomous driving vehicles based on group behavior prediction. Background Art

[0002] With the development of autonomous driving technology, the intelligence and automation level of vehicles are constantly improving, which provides a technical basis for group vehicle transportation. Autonomous driving vehicles can perceive and make decisions about the environment through sensors and algorithms, thereby achieving more efficient and safe transportation in group transportation.

[0003] Traditional pick-up and drop-off systems often rely on fixed routes and schedules, lacking flexibility and real-time performance. This leads to a waste of resources and increased waiting time for passengers. In addition, these systems are usually unable to effectively respond to emergencies such as traffic congestion or weather changes, and cannot predict unexpected traffic situations in advance.

[0004] At present, the focus of research on traffic flow distribution tends to be on the mechanism of path selection under different information conditions or more effective solution algorithms, but there is less research on its important foundation, "dynamic decision-making needs of group travel", which limits the progress of dynamic traffic distribution theory and application research, resulting in the existing path planning methods being difficult to reflect group dynamic decision-making and unable to meet people's needs in scientific travel, intelligent guidance, etc. Summary of the invention

[0005] The purpose of the present invention is to provide an automatic driving vehicle group pick-up optimization system based on group behavior prediction to solve the above-mentioned problems existing in the prior art.

[0006] The specific application is as follows:

[0007] The group pick-up optimization system for autonomous vehicles based on group behavior prediction includes a passenger group data collection module, a group behavior prediction and pattern analysis module, a multimodal large model optimization module, a group pick-up point priority sorting and dynamic adjustment module, and a real-time group interaction and feedback module;

[0008] The passenger group data collection module is used to collect real-time data of each passenger in the group, including their current location information, scheduled pick-up point, demand preferences, and historical behavior data;

[0009] The group behavior prediction and pattern analysis module analyzes the behavior pattern of each passenger in the group through a multimodal large model, combines current environmental data and historical behavior, and predicts the future action path and time schedule of each passenger in the group;

[0010] The multimodal large model optimization module is used to combine the group behavior prediction results of the group behavior prediction and pattern analysis module, integrate various input data through the multimodal large model, and dynamically generate the optimal pick-up point and route planning;

[0011] The group pick-up point priority sorting and dynamic adjustment module dynamically adjusts the pick-up order and pick-up point priority by predicting the needs and behaviors of passengers in the group;

[0012] The real-time group interaction and feedback module maintains real-time interaction with each passenger through mobile devices, providing dynamic updates of current pick-up points and driving routes, as well as feedback on the group pick-up sequence;

[0013] The passenger group data collection module, the group behavior prediction and pattern analysis module, the multimodal large model optimization module, the group pick-up point priority sorting and dynamic adjustment module and the real-time group interaction and feedback module are interconnected.

[0014] Furthermore, the passenger group data collection module includes a location data collection submodule, a demand input submodule and a behavior data integration submodule;

[0015] The location data acquisition submodule obtains the current location of the passenger through GPS, Wi-Fi signal, and Bluetooth;

[0016] The demand input submodule is used for passengers to input pick-up and drop-off requirements through mobile devices, including specific pick-up and drop-off points, time limits, and preferred routes;

[0017] The behavior data integration submodule performs personalized analysis based on the passenger's historical behavior data, which includes riding habits, frequently used pick-up and drop-off points, and preferred driving routes.

[0018] Furthermore, the specific implementation process of the position data collection submodule is as follows:

[0019] L1: Data collection, obtaining outdoor location information of passengers through GPS and generating GPS data; building a fingerprint database by collecting Wi-Fi signal characteristics for indoor location information of passengers and generating Wi-Fi data; collecting beacon broadcast data packets received by passengers' terminal devices through Bluetooth beacons to obtain location information and generate BLE data, wherein the beacon broadcast data packets at least include MAC address and RSSI signal strength;

[0020] L2: Preprocessing, preprocessing the collected GPS data, Wi-Fi data and BLE data, including noise filtering and data synchronization;

[0021] L3: Data fusion, which integrates GPS data, Wi-Fi data, and BLE data through the Kalman filter algorithm;

[0022] The specific implementation process of L3 is as follows:

[0023] L31: Set the initial state and covariance matrix of the Kalman filter, including the process noise covariance and the observation noise covariance;

[0024] L32: Predict the current state and covariance based on the previous state;

[0025] L33: Update state estimates and covariances using new observations;

[0026] L34: Repeat steps L32 and L33 to achieve continuous tracking and optimization of passenger positions;

[0027] The specific implementation process of the demand input submodule is as follows:

[0028] Passengers submit pick-up requests through mobile devices. Natural language processing technology is used to process the pick-up requests submitted by passengers. The pre-trained BERT language model is used to analyze the demand description in the pick-up request of the passenger, and the demand description is converted into a pick-up demand. The named entity recognition (NER) technology is used to identify the geographic location and time in the pick-up demand, and the geographic location and time are converted into parameters in the pick-up plan.

[0029] The specific implementation process of the behavior data integration submodule is as follows:

[0030] Using the historical behavior data of passengers, a hidden Markov model is used to model the passenger behavior and generate a prediction model for the passenger's future behavior;

[0031] The input of the passenger future behavior prediction model training is the passenger's historical behavior data, and the output is the passenger's future behavior pattern, which includes whether the passenger is expected to show up on time, the passenger's preferred route, and the passenger's common behavior in a specific time period;

[0032] The hidden Markov model formula is:

[0033] P(A|θ)=∑ allstates P(A,Q|θ),

[0034] Among them, A is the observed behavior sequence, θ is the model parameter, which is obtained by maximum likelihood estimation, and Q is the state of the behavior sequence.

[0035] Further, the group behavior prediction and pattern analysis module includes a behavior pattern analysis submodule, a trajectory prediction submodule and a group behavior prediction submodule;

[0036] The behavior pattern analysis submodule identifies the behavior pattern of the passenger by analyzing the historical behavior data of the passenger, wherein the behavior pattern at least includes a preference for getting on and off the bus at certain specific times and locations;

[0037] The trajectory prediction submodule predicts the passenger's future movement path based on the current location and environmental information;

[0038] The group behavior prediction submodule generates the best pick-up and drop-off sequence and path for the entire group based on the behavior patterns of multiple passengers, and adjusts the plan according to the needs of different passengers.

[0039] Furthermore, the specific implementation process of the behavior pattern analysis submodule is as follows:

[0040] Based on the historical behavior data of passengers, the clustering algorithm K-means is used to analyze the behavior patterns of different passengers, and classification is performed based on the results of the pattern analysis;

[0041] The specific implementation process of the trajectory prediction submodule is as follows:

[0042] The passenger's movement trajectory is predicted using a spatiotemporal trajectory prediction model based on a long short-term memory network (LSTM), wherein the training input of the spatiotemporal trajectory prediction model is the passenger's historical trajectory, and the output is the passenger's future movement trajectory;

[0043] The specific implementation process of the group behavior prediction submodule is as follows:

[0044] A model combining graph convolutional network (GCN) and Transformer attention mechanism is used to globally optimize the behaviors of multiple passengers. The graph convolutional network (GCN) predicts the overall behavior pattern by processing the relationship network between group members, and Transformer is used to capture the complex dependencies of the relationship network to generate the optimal pick-up and drop-off sequence and path plan.

[0045] The specific implementation process of the optimal pick-up sequence and path plan is as follows:

[0046] S1: Multiple passenger service area division based on K-Means++ algorithm;

[0047] S2: Determination of multiple passenger sharing points based on DBSCAN algorithm;

[0048] S3: Generation of optimal pick-up and drop-off sequence paths based on maximum likelihood estimation;

[0049] The specific implementation process of S1 is as follows:

[0050] S11: Input a coordinate data set N containing n passenger demand points, N = {(xi ,y i )|i=1,2,3,...,n}, input the preset category number k, where the value of k is determined according to the required quantity, that is, Q is the vehicle seating capacity;

[0051] S12: Randomly select a point (x p ,y p ) as the cluster center, and calculate the shortest distance between each sample and the current cluster center using D pi Indicates that the D pi Use the third-party map API interface to obtain the longitude and latitude coordinates of the passenger demand points, and then obtain the driving distance between the two passenger demand points;

[0052] S13: Calculate the probability of each sample point being selected as the next cluster center point

[0053] S14: Repeat steps S12 and S13 until k cluster centers are selected;

[0054] S15: Calculate the distances between the k cluster centers of each data sample, and select the cluster center with the closest distance to be classified into one category;

[0055] S16: Calculate the mean distance in each cluster, update the cluster center by the mean distance between the horizontal coordinate and the vertical coordinate, and repeat steps S15 and S16 until convergence;

[0056] S17: Extract clustering results, where the data sample size in each cluster is no greater than the clustering result of the passenger capacity of a single vehicle, as the single n passenger demand points;

[0057] The specific implementation process of S2 is:

[0058] S21: Input data set N = {(x i ,y i )|i=1,2,3,...,N}, specify the radius of Eps neighborhood as r, the minimum number of samples in MinPts neighborhood as q, where MinPts and Eps are both parameters in DBSCAN clustering algorithm;

[0059] S22: Initialize the number of clusters k'=0, initialize the unvisited sample set P=D, and the multiplication point set

[0060] S23: Randomly select a data sample o i =(x i ,y i ), if |N Eps (o i )|≥MinPts, then oi Is the core object, o i Add Cluster C i In the collection, C i ={o i}, and the neighborhood subset N Eps (o i ) added to C i middle;

[0061] S24: Update data sample P=PC i Repeat step S23 until each data sample point is visited and the final multiplication point set C is obtained. k' ={C1,C2,...,C i};

[0062] S25: For the same cluster C i A demand point in the random selection is used as the final boarding / disembarking point;

[0063] The specific implementation process of S3 is as follows:

[0064] The method of univariate linear regression is adopted to fit all passenger pool points through the least squares method to generate a uniform linear regression line of the optimal pick-up and drop-off sequence. Then, the linear regression line is input into the actual road network map to calculate the intersection points through the Dijkstra algorithm. Finally, the shortest path with cluster P0 as the starting point is generated, and each node is marked as P1, P2, ..., P n .

[0065] Further, the multimodal large model optimization module includes a multi-data input processing submodule, a dynamic traffic analysis submodule and a personalized optimization submodule;

[0066] The multi-data input processing submodule analyzes the instructions and requirements input by the passenger through natural language processing technology, and confirms the pick-up point location in combination with the image and text data;

[0067] The dynamic traffic analysis submodule dynamically optimizes the pick-up point selection and driving route by analyzing real-time traffic data, weather information and road conditions;

[0068] The personalized optimization submodule generates a personalized pick-up and drop-off plan for each passenger, and the personalized pick-up and drop-off plan needs to be combined with the specific needs of the passenger and the needs of other members in the group.

[0069] Furthermore, the specific implementation process of the multiple data input processing submodule is as follows:

[0070] Using a multimodal fusion CLIP model to process the input of passenger voice, image and text, the multimodal fusion CLIP model outputs the passenger's location information, and the passenger's location information is used to fuse the real-time image data to confirm the pick-up point;

[0071] The specific implementation process of the dynamic traffic analysis submodule is as follows:

[0072] Through the deep Q network model and proximal operator Proximal, real-time traffic information and road conditions are analyzed to dynamically adjust the vehicle pick-up and drop-off routes;

[0073] The specific implementation process of the personalized optimization submodule is as follows:

[0074] A genetic algorithm for multi-objective optimization is used to generate an optimization solution for each passenger's personalized needs, including preferred pick-up and drop-off times and specific route preferences.

[0075] Further, the group pick-up point priority sorting and dynamic adjustment module includes a pick-up point sorting submodule and a dynamic adjustment submodule;

[0076] The pick-up point sorting submodule prioritizes the pick-up points according to the needs of each passenger, wherein the needs include at least time limits and preferred locations;

[0077] The dynamic adjustment submodule is used to automatically adjust the pick-up order and route when the needs or position of a passenger in the group changes, and the position change at least includes late arrival or temporary change of the pick-up point.

[0078] Furthermore, the specific implementation process of prioritizing the pick-up and drop-off points in the pick-up and drop-off point sorting submodule is as follows:

[0079] Priority sorting uses a multi-criteria decision-making TOPSIS algorithm to generate a priority sorting of pick-up points based on passenger needs, calculate the positive ideal solution and negative ideal solution of each pick-up point, and generate a heat map based on the positive ideal solution and negative ideal solution. According to the distribution of each pick-up point in the heat map, dense distribution has a high score, and sparse distribution has a low score. The best pick-up order is determined based on the scores. The passenger's needs include time constraints and location preferences;

[0080] The specific implementation process of the dynamic adjustment submodule is as follows:

[0081] The pickup sequence is adjusted in real time using the reinforcement learning algorithm DQN. When passenger demand or location changes, the pickup sequence and path are recalculated based on the current environment, and the plan is adjusted based on the latest traffic and passenger data. The current environment includes passenger lateness and temporary changes to the pickup point.

[0082] Further, the real-time group interaction and feedback module includes a real-time notification submodule and a feedback and adjustment submodule;

[0083] The real-time notification submodule is used to notify each passenger of the update of the pick-up point, estimated arrival time, and vehicle location through the application APP, and allow passengers to view the pick-up order and route in the group in real time;

[0084] The feedback and adjustment submodule is used for passengers to provide real-time feedback or adjust their pick-up and drop-off requirements through the APP, and dynamically adjust the group pick-up and drop-off plan based on the feedback or adjustment of the pick-up and drop-off requirements.

[0085] Compared with the prior art, the embodiments of the present invention achieve the following beneficial effects:

[0086] The embodiment of the present invention provides a passenger group data collection module for collecting real-time data of each passenger in the group, including their current location information, scheduled pick-up and drop-off points, demand preferences, and historical behavior data; the group behavior prediction and pattern analysis module analyzes the behavior pattern of each passenger in the group through a multimodal large model, and combines the current environmental data and historical behavior to predict the future action path and time schedule of each passenger in the group; the multimodal large model optimization module is used to combine the group behavior prediction results of the group behavior prediction and pattern analysis module, integrate various input data through the multimodal large model, and dynamically generate the optimal pick-up and drop-off points and path planning; group pick-up and drop-off The point priority sorting and dynamic adjustment module dynamically adjusts the pick-up order and pick-up point priority by predicting the needs and behaviors of passengers in the group; the real-time group interaction and feedback module maintains real-time interaction with each passenger through mobile devices, provides dynamic updates of current pick-up points and driving routes, and feedback on the group pick-up order; the present invention conducts real-time analysis of the behaviors, needs, locations and future movement trajectories of multiple passengers, combined with group behavior prediction technology, so that the autonomous driving vehicle can dynamically adjust the pick-up points and driving routes to maximize the pick-up efficiency and passenger experience. The system of the present invention is particularly suitable for shared travel, multi-passenger carpooling, and dynamically changing multi-person pick-up scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0087] Figure 1 is a system architecture diagram of an automatic driving vehicle group pick-up and drop-off optimization system based on group behavior prediction provided by an embodiment of the present invention;

[0088] Figure 2 This is a diagram of a passenger group data collection module of an automatic driving vehicle group pick-up and drop-off optimization system based on group behavior prediction provided by an embodiment of the present invention;

[0089] Figure 3 It is a group behavior prediction and pattern analysis module diagram of an automatic driving vehicle group pick-up optimization system based on group behavior prediction provided by an embodiment of the present invention;

[0090] Figure 4 It is a multi-modal large model optimization module diagram of an automatic driving vehicle group pick-up optimization system based on group behavior prediction provided by an embodiment of the present invention;

[0091] Figure 5 It is a group pick-up point priority sorting and dynamic adjustment module diagram of the autonomous driving vehicle group pick-up optimization system based on group behavior prediction provided by an embodiment of the present invention;

[0092] Figure 6 It is a real-time group interaction and feedback module diagram of an autonomous driving vehicle group pick-up optimization system based on group behavior prediction provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0093] The present invention will be described in detail below in conjunction with the accompanying drawings.

[0094] Example 1

[0095] The embodiment of the present invention provides an optimization system for group pick-up and drop-off of autonomous driving vehicles based on group behavior prediction, such as Figure 1 As shown, it includes a passenger group data collection module, a group behavior prediction and pattern analysis module, a multimodal large model optimization module, a group pick-up point priority sorting and dynamic adjustment module, and a real-time group interaction and feedback module;

[0096] The passenger group data collection module is used to collect real-time data of each passenger in the group, including their current location information, scheduled pick-up point, demand preferences, and historical behavior data;

[0097] The group behavior prediction and pattern analysis module analyzes the behavior pattern of each passenger in the group through a multimodal large model, combines current environmental data and historical behavior, and predicts the future action path and time schedule of each passenger in the group;

[0098] The multimodal large model optimization module is used to combine the group behavior prediction results of the group behavior prediction and pattern analysis module, integrate various input data through the multimodal large model, and dynamically generate the optimal pick-up point and route planning;

[0099] The group pick-up point priority sorting and dynamic adjustment module dynamically adjusts the pick-up order and pick-up point priority by predicting the needs and behaviors of passengers in the group;

[0100] The real-time group interaction and feedback module maintains real-time interaction with each passenger through mobile devices, providing dynamic updates of current pick-up points and driving routes, as well as feedback on the group pick-up sequence;

[0101] The passenger group data collection module, the group behavior prediction and pattern analysis module, the multimodal large model optimization module, the group pick-up point priority sorting and dynamic adjustment module and the real-time group interaction and feedback module are interconnected.

[0102] Furthermore, the passenger group data collection module includes a location data collection submodule, a demand input submodule and a behavior data integration submodule, see Figure 2 ;

[0103] The location data acquisition submodule obtains the current location of the passenger through GPS, Wi-Fi signal, and Bluetooth; the demand input submodule is used for passengers to input pick-up and drop-off requirements through mobile devices, including specific pick-up and drop-off points, time limits, and preferred routes;

[0104] The behavior data integration submodule performs personalized analysis based on the passenger's historical behavior data, which includes riding habits, frequently used pick-up and drop-off points, and preferred driving routes.

[0105] Furthermore, the specific implementation process of the position data collection submodule is as follows:

[0106] L1: Data collection, obtaining outdoor location information of passengers through GPS and generating GPS data; building a fingerprint database by collecting Wi-Fi signal characteristics for indoor location information of passengers and generating Wi-Fi data; collecting beacon broadcast data packets received by passengers' terminal devices through Bluetooth beacons to obtain location information and generate BLE data, wherein the beacon broadcast data packets at least include MAC address and RSSI signal strength;

[0107] L2: Preprocessing, preprocessing the collected GPS data, Wi-Fi data and BLE data, including noise filtering and data synchronization;

[0108] L3: Data fusion, which integrates GPS data, Wi-Fi data, and BLE data through the Kalman filter algorithm;

[0109] The specific implementation process of L3 is as follows:

[0110] L31: Set the initial state and covariance matrix of the Kalman filter, including the process noise covariance and the observation noise covariance;

[0111] L32: Predict the current state and covariance based on the previous state;

[0112] L33: Update state estimates and covariances using new observations;

[0113] L34: Repeat steps L32 and L33 to achieve continuous tracking and optimization of passenger positions;

[0114] The specific implementation process of the demand input submodule is as follows:

[0115] Passengers submit pick-up requests through mobile devices. Natural language processing technology is used to process the pick-up requests submitted by passengers. The pre-trained BERT language model is used to analyze the demand description in the pick-up request of the passenger, and the demand description is converted into a pick-up demand. The named entity recognition (NER) technology is used to identify the geographic location and time in the pick-up demand, and the geographic location and time are converted into parameters in the pick-up plan.

[0116] The specific implementation process of the behavior data integration submodule is as follows:

[0117] Using the historical behavior data of passengers, a hidden Markov model is used to model the passenger behavior and generate a prediction model for the passenger's future behavior;

[0118] The input of the passenger future behavior prediction model training is the passenger's historical behavior data, and the output is the passenger's future behavior pattern, which includes whether the passenger is expected to show up on time, the passenger's preferred route, and the passenger's common behavior in a specific time period;

[0119] The hidden Markov model formula is:

[0120] P(A|θ)=∑ allstates P(A,Q|θ),

[0121] Among them, A is the observed behavior sequence, θ is the model parameter, which is obtained by maximum likelihood estimation, and Q is the state of the behavior sequence.

[0122] Furthermore, the group behavior prediction and pattern analysis module includes a behavior pattern analysis submodule, a trajectory prediction submodule and a group behavior prediction submodule, see Figure 3 ;

[0123] The behavior pattern analysis submodule identifies the behavior pattern of the passenger by analyzing the historical behavior data of the passenger, wherein the behavior pattern at least includes a preference for getting on and off the bus at certain specific times and locations;

[0124] The trajectory prediction submodule predicts the passenger's future movement path based on the current location and environmental information;

[0125] The group behavior prediction submodule generates the best pick-up and drop-off sequence and path for the entire group based on the behavior patterns of multiple passengers, and adjusts the plan according to the needs of different passengers.

[0126] Specifically, the passenger group data collection module and the group behavior prediction and pattern analysis module interact in the following process:

[0127] Interaction Description:

[0128] Data transmission: The passenger group data collection module first obtains each passenger's current location information, historical behavior data (such as the time and place of frequent rides), and demand input (such as preferred pick-up points and route selection). These data are transmitted to the group behavior prediction and pattern analysis module in real time;

[0129] Real-time update: The data collection module passes the latest status of the data to the behavior prediction module based on the changes in the passenger's behavior, such as when the passenger temporarily changes the pick-up and drop-off requirements or the location information changes, to trigger the recalculation of the behavior pattern and trajectory prediction;

[0130] Interaction algorithms and models:

[0131] The Markov model HMM is combined with the long short-term memory network LSTM:

[0132] The Hidden Markov Model analyzes passengers’ long-term behavior patterns based on historical behavior data, including common pick-up and drop-off locations and frequencies.

[0133] The LSTM model performs spatiotemporal predictions in the short term, predicting the future course of action of passengers by analyzing their current location and historical trajectory;

[0134] Data Fusion:

[0135] Through the complementary combination of LSTM, short-term trajectories can be predicted based on sequence data;

[0136] Dynamic data update: The data acquisition module updates the passenger location information in real time and inputs it into the LSTM network. The updated location information triggers the trajectory prediction module to re-evaluate the future path.

[0137] Furthermore, the specific implementation process of the behavior pattern analysis submodule is as follows:

[0138] Based on the historical behavior data of passengers, the clustering algorithm K-means is used to analyze the behavior patterns of different passengers, and classification is performed based on the results of the pattern analysis;

[0139] The specific implementation process of the trajectory prediction submodule is as follows:

[0140] The passenger's movement trajectory is predicted using a spatiotemporal trajectory prediction model based on a long short-term memory network (LSTM), wherein the training input of the spatiotemporal trajectory prediction model is the passenger's historical trajectory, and the output is the passenger's future movement trajectory;

[0141] The specific implementation process of the group behavior prediction submodule is as follows:

[0142] A model combining graph convolutional network (GCN) and Transformer attention mechanism is used to globally optimize the behaviors of multiple passengers. The graph convolutional network (GCN) predicts the overall behavior pattern by processing the relationship network between group members, and Transformer is used to capture the complex dependencies of the relationship network to generate the optimal pick-up and drop-off sequence and path plan.

[0143] The specific implementation process of the optimal pick-up sequence and path plan is as follows:

[0144] S1: Multiple passenger service area division based on K-Means++ algorithm;

[0145] S2: Determination of multiple passenger sharing points based on DBSCAN algorithm;

[0146] S3: Generation of optimal pick-up and drop-off sequence paths based on maximum likelihood estimation;

[0147] The specific implementation process of S1 is as follows:

[0148] S11: Input a coordinate data set N containing n passenger demand points, N = {(x i ,y i )|i=1,2,3,...,n}, input the preset category number k, where the value of k is determined according to the required quantity, that is, Q is the vehicle seating capacity;

[0149] S12: Randomly select a point (x p ,y p ) as the cluster center, and calculate the shortest distance between each sample and the current cluster center using D pi Indicates that the D pi Use the third-party map API interface to obtain the longitude and latitude coordinates of the passenger demand points, and then obtain the driving distance between the two passenger demand points;

[0150] S13: Calculate the probability of each sample point being selected as the next cluster center point

[0151] S14: Repeat steps S12 and S13 until k cluster centers are selected;

[0152] S15: Calculate the distances between the k cluster centers of each data sample, and select the cluster center with the closest distance to be classified into one category;

[0153] S16: Calculate the mean distance in each cluster, update the cluster center by the mean distance between the horizontal coordinate and the vertical coordinate, and repeat steps S15 and S16 until convergence;

[0154] S17: Extract clustering results, where the data sample size in each cluster is no greater than the clustering result of the passenger capacity of a single vehicle, as the single n passenger demand points;

[0155] The specific implementation process of S2 is:

[0156] S21: Input data set N = {(x i ,y i )|i=1,2,3,...,N}, specify the radius of Eps neighborhood as r, the minimum number of samples in MinPts neighborhood as q, where MinPts and Eps are both parameters in DBSCAN clustering algorithm;

[0157] S22: Initialize the number of clusters k'=0, initialize the unvisited sample set P=D, and the multiplication point set

[0158] S23: Randomly select a data sample o i =(x i ,y i ), if |N Eps (o i )|≥MinPts, then o i Is the core object, o i Add Cluster C i In the collection, C i ={o i}, and the neighborhood subset N Eps (o i ) added to C i middle;

[0159] S24: Update data sample P=PC i Repeat step S23 until each data sample point is visited and the final multiplication point set C is obtained. k' ={C1,C2,...,C i};

[0160] S25: For the same cluster C i A demand point in the random selection is used as the final boarding / disembarking point;

[0161] The specific implementation process of S3 is as follows:

[0162] The method of univariate linear regression is adopted to fit all passenger pool points through the least squares method to generate a uniform linear regression line of the optimal pick-up and drop-off sequence. Then, the linear regression line is input into the actual road network map to calculate the intersection points through the Dijkstra algorithm. Finally, the shortest path with cluster P0 as the starting point is generated, and each node is marked as P1, P2, ..., P n .

[0163] Furthermore, the multi-modal large model optimization module includes a multi-data input processing submodule, a dynamic traffic analysis submodule and a personalized optimization submodule, please refer to Figure 4 ;

[0164] The multi-data input processing submodule analyzes the instructions and requirements input by the passenger through natural language processing technology, and confirms the pick-up point location in combination with the image and text data;

[0165] The dynamic traffic analysis submodule dynamically optimizes the pick-up point selection and driving route by analyzing real-time traffic data, weather information and road conditions;

[0166] The personalized optimization submodule generates a personalized pick-up and drop-off plan for each passenger, and the personalized pick-up and drop-off plan needs to be combined with the specific needs of the passenger and the needs of other members in the group.

[0167] Specifically, the interaction process between the group behavior prediction and pattern analysis module and the multimodal large model optimization module is as follows:

[0168] Interaction Description:

[0169] Input data: The group behavior prediction and pattern analysis module provides passengers' future trajectory predictions, behavior pattern analysis results, and preliminary route planning based on traffic conditions. These data are used by the multimodal large model optimization module to comprehensively optimize pick-up points and routes.

[0170] Environmental perception: The multimodal large model module combines real-time information such as traffic conditions, weather, and road occlusions to further optimize the results of group behavior prediction, including adjusting routes to avoid congestion or delays;

[0171] Interaction algorithms and models:

[0172] Combining graph convolutional network GCN with Transformer:

[0173] GCN is used to model the relationship network between passengers, including their geographical locations, the closeness of demand time, etc., to generate preliminary group pick-up and drop-off sequencing;

[0174] The Transformer attention mechanism helps capture the complex dependencies between group members and optimize the final pick-up points and routes;

[0175] Group behavior prediction generates a graph structure between groups through the graph convolutional network GCN;

[0176] Traffic optimization and dynamic environment analysis: The multimodal large model integrates the results of the group behavior prediction module and applies a deep reinforcement learning model. The deep reinforcement learning model includes but is not limited to DQN or PPO for path planning. By continuously exploring the interaction with the environment's traffic and weather, the deep reinforcement learning model will adjust the pick-up path.

[0177] Furthermore, the specific implementation process of the multiple data input processing submodule is as follows:

[0178] Using a multimodal fusion CLIP model to process the input of passenger voice, image and text, the multimodal fusion CLIP model outputs the passenger's location information, and the passenger's location information is used to fuse the real-time image data to confirm the pick-up point;

[0179] The specific implementation process of the dynamic traffic analysis submodule is as follows:

[0180] Through the deep Q network model and proximal operator Proximal, real-time traffic information and road conditions are analyzed to dynamically adjust the vehicle pick-up and drop-off routes;

[0181] The specific implementation process of the personalized optimization submodule is as follows:

[0182] A genetic algorithm for multi-objective optimization is used to generate an optimization solution for each passenger's personalized needs, including preferred pick-up and drop-off times and specific route preferences.

[0183] Furthermore, the group pick-up point priority sorting and dynamic adjustment module includes a pick-up point sorting submodule and a dynamic adjustment submodule, see Figure 5 ;

[0184] The pick-up point sorting submodule prioritizes the pick-up points according to the needs of each passenger, wherein the needs include at least time limits and preferred locations;

[0185] The dynamic adjustment submodule is used to automatically adjust the pick-up order and route when the needs or position of a passenger in the group changes, and the position change at least includes late arrival or temporary change of the pick-up point.

[0186] Specifically, the multimodal large model optimization module and the group pick-up point priority sorting and dynamic adjustment module interact in the following process:

[0187] Interaction Description:

[0188] Optimize data transmission: The multimodal large model module will pass the initially optimized pick-up points and route plans to the pick-up point priority sorting and dynamic adjustment module, which will dynamically adjust the priority of the pick-up points based on the passenger's priority needs, real-time changes in passenger needs or locations.

[0189] Real-time adjustment: The priority sorting module is updated based on real-time data, and the adjusted pick-up and drop-off point sequence is fed back to the multimodal large model, triggering re-optimization of the path;

[0190] Interaction algorithms and models:

[0191] Combining TOPSIS with deep reinforcement learning:

[0192] When determining the order of pick-up and drop-off points, the TOPSIS multi-criteria decision model performs ordering and generates a series of priorities based on the needs of passengers, including time and location preferences for pick-up and drop-off points.

[0193] The reinforcement learning model DQN is used to adjust the order of pick-up and drop-off points in real time. When passenger demand or the environment changes, the reinforcement learning model DQN readjusts the pick-up and drop-off points and routes based on the new feedback;

[0194] The formula of the TOPSIS multi-criteria decision model is:

[0195]

[0196] Among them, S i Score the priority of the pick-up point, w j is the weight of each factor, a ij is the score of each pick-up point under the jth criterion;

[0197] Dynamic priority adjustment formula:

[0198] P(t+1)=P(t)+αΔP(t), where P(t+1) represents the updated pick-up point priority, α represents the learning rate, and ΔP(t) represents the priority adjustment amount caused by real-time feedback.

[0199] Furthermore, the specific implementation process of prioritizing the pick-up and drop-off points in the pick-up and drop-off point sorting submodule is as follows:

[0200] Priority sorting uses a multi-criteria decision-making TOPSIS algorithm to generate a priority sorting of pick-up points based on passenger needs, calculate the positive ideal solution and negative ideal solution of each pick-up point, and generate a heat map based on the positive ideal solution and negative ideal solution. According to the distribution of each pick-up point in the heat map, dense distribution has a high score, and sparse distribution has a low score. The best pick-up order is determined based on the scores. The passenger's needs include time constraints and location preferences;

[0201] The specific implementation process of the dynamic adjustment submodule is as follows:

[0202] The pickup sequence is adjusted in real time using the reinforcement learning algorithm DQN. When passenger demand or location changes, the pickup sequence and path are recalculated based on the current environment, and the plan is adjusted based on the latest traffic and passenger data. The current environment includes passenger lateness and temporary changes to the pickup point.

[0203] Furthermore, the real-time group interaction and feedback module includes a real-time notification submodule and a feedback and adjustment submodule, see Figure 6 ;

[0204] Specifically, the group pick-up point priority sorting and dynamic adjustment module interacts with the real-time group interaction and feedback module in the following process:

[0205] Interaction Description:

[0206] Interactive data: The priority sorting and dynamic adjustment module pushes the adjusted pick-up order and route to the real-time interaction and feedback module, notifying each passenger of the new pick-up plan and the vehicle's dynamic itinerary information in real time;

[0207] Feedback mechanism: Passengers can use the real-time interaction module to provide feedback on changes in their needs, including temporary changes to the pick-up and drop-off points. This feedback information will be passed to the priority sorting and dynamic adjustment module through the interaction module, and further fed back to the multimodal large model module to trigger the system's re-optimization process;

[0208] Interaction algorithms and models:

[0209] Real-time communication mechanism and intelligent feedback:

[0210] By ensuring low-latency real-time communication through WebSocket and MQTT protocols, the system can quickly push updated information to passengers when the pick-up and drop-off plan is adjusted.

[0211] The intelligent dialogue system Dialogflow receives feedback from passengers, analyzes their instructions through NLP, and converts them into instructions that the system can understand, further triggering data and algorithm linkage between modules within the system;

[0212] Real-time feedback formula:

[0213] T total =T notification +T response +T ajustment ,

[0214] Among them, T total represents the total time of passenger feedback response, T notification Indicates the notification time, T responserepresents the passenger response time, T ajustment Indicates the time for the system to adjust the pick-up and drop-off sequence.

[0215] The real-time notification submodule is used to notify each passenger of the update of the pick-up point, estimated arrival time, and vehicle location through the application APP, and allow passengers to view the pick-up order and route in the group in real time;

[0216] The feedback and adjustment submodule is used for passengers to provide real-time feedback or adjust their pick-up and drop-off requirements through the APP, and dynamically adjust the group pick-up and drop-off plan based on the feedback or adjustment of the pick-up and drop-off requirements.

[0217] The specific system comprehensive optimization process is as follows:

[0218] 1. Data collection: Passengers’ location information, needs, and behavior patterns are input into the system in real time through the data collection module;

[0219] 2. Behavior prediction: The group behavior prediction module predicts the trajectory of passengers based on the collected data and preliminarily generates the pick-up and drop-off sequence and route planning;

[0220] 3. Multimodal optimization: The multimodal large model optimization module combines real-time traffic, weather, and road information to dynamically adjust the initial path planning;

[0221] 4. Priority sorting and dynamic adjustment: Dynamically adjust the pick-up point priority and route sequence based on the changes in passenger demand and behavior prediction results;

[0222] 5. Real-time feedback and updates: The system pushes the latest pick-up and drop-off plans to passengers through the real-time interaction module, and passenger feedback further triggers system re-optimization.

[0223] Example 2: Group Optimization in Multi-Passenger Carpooling

[0224] 1. Background description: An autonomous taxi receives carpooling requests from multiple passengers. The system needs to dynamically optimize the pick-up and drop-off order based on the passengers’ locations, needs, and time constraints.

[0225] 2. Group data collection:

[0226] Each passenger inputs their pick-up and drop-off requirements through their smartphone and uploads their current location information;

[0227] The system confirms the specific location of each passenger through GPS and Wi-Fi signals, and integrates their historical behavior data to predict the passenger's future movement trajectory;

[0228] 3. Group behavior prediction and optimization: The system analyzes the behavior patterns of each passenger in the group and generates the optimal pick-up sequence and route planning based on traffic conditions;

[0229] 4. Dynamic adjustment: If a passenger temporarily changes the pick-up point, the system recalculates the pick-up order in the group and pushes the updated plan to all passengers.

[0230] Example 3: Prioritization of Pickup and Dropoff Points in a Ride Sharing System

[0231] 1. Background description: Four passengers share a ride during peak hours. The system needs to generate the best pick-up order and driving route based on their pick-up needs and time constraints;

[0232] 2. Data input:

[0233] The system receives passengers' scheduled pick-up and drop-off points and time limits, and combines their historical behavior data for personalized optimization. Using GPS and real-time traffic data, the system calculates the optimal pick-up and drop-off sequence to ensure that time-sensitive passengers are picked up first;

[0234] 3. Group behavior prediction and feedback: The system pushes real-time updates of pick-up points and estimated arrival times to passengers, and allows passengers to adjust pick-up points according to their needs;

[0235] 4. Dynamic Optimization: In case of road closures or traffic congestion, the system automatically adjusts the pick-up order and notifies all passengers of the new route and estimated arrival time.

[0236] The algorithm and display provided herein are not inherently related to any particular computer, virtual system or other device. Various general purpose systems can also be used together with the teachings based on this. According to the above description, it is obvious that the structure required for constructing such systems. In addition, the present invention is not directed to any specific programming language either. It should be understood that various programming languages ​​can be utilized to realize the content of the present invention described herein, and the description of the above specific languages ​​is for disclosing the best mode of the present invention.

[0237] In the description provided herein, a large number of specific details are described. However, it is understood that embodiments of the present invention can be practiced without these specific details. In some instances, well-known methods, structures and techniques are not shown in detail so as not to obscure the understanding of this description.

[0238] Similarly, it should be understood that in order to streamline the present disclosure and aid in understanding one or more of the various inventive aspects, in the above description of exemplary embodiments of the present invention, various features of the present invention are sometimes grouped together into a single embodiment, figure, or description thereof. However, this disclosed method should not be interpreted as reflecting the intention that the claimed invention requires more features than those explicitly recited in each claim. More specifically, as reflected in the claims below, inventive aspects lie in less than all the features of the individual embodiments disclosed above. Therefore, the claims that follow the specific embodiment are hereby expressly incorporated into the specific embodiment, with each claim itself serving as a separate embodiment of the present invention.

[0239] Those skilled in the art will appreciate that the modules in the devices in the embodiments may be adaptively changed and arranged in one or more devices different from the embodiments. The modules or units or components in the embodiments may be combined into one module or unit or component, and in addition they may be divided into a plurality of submodules or subunits or subcomponents. Except that at least some of such features and / or processes or units are mutually exclusive, all features disclosed in this specification (including the accompanying claims, abstracts and drawings) and all processes or units of any method or device disclosed in this manner may be combined in any combination. Unless otherwise expressly stated, each feature disclosed in this specification (including the accompanying claims, abstracts and drawings) may be replaced by an alternative feature providing the same, equivalent or similar purpose.

[0240] In addition, those skilled in the art will appreciate that, although some embodiments herein include certain features included in other embodiments but not other features, the combination of features of different embodiments is meant to be within the scope of the present invention and form different embodiments. For example, in the claims below, any one of the claimed embodiments may be used in any combination.

[0241] The various component embodiments of the present invention may be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. It should be understood by those skilled in the art that a microprocessor or digital signal processor (DSP) may be used in practice to implement some or all of the functions of some or all of the components in the apparatus according to an embodiment of the present invention. The present invention may also be implemented as a device or device program (e.g., a computer program and a computer program product) for executing part or all of the methods described herein. Such a program implementing the present invention may be stored on a computer-readable medium, or may have the form of one or more signals. Such a signal may be downloaded from an Internet website, or provided on a carrier signal, or provided in any other form.

Claims

1. An automated driving vehicle group pick-up and drop-off optimization system based on group behavior prediction, characterized in that: It includes passenger group data collection module, group behavior prediction and pattern analysis module, multimodal large model optimization module, group pick-up point priority sorting and dynamic adjustment module and real-time group interaction and feedback module; The passenger group data collection module is used to collect real-time data of each passenger in the group, including their current location information, scheduled pick-up point, demand preferences, and historical behavior data; The group behavior prediction and pattern analysis module analyzes the behavior pattern of each passenger in the group through a multimodal large model, combines current environmental data and historical behavior, and predicts the future action path and time schedule of each passenger in the group; The multimodal large model optimization module is used to combine the group behavior prediction results of the group behavior prediction and pattern analysis module, integrate various input data through the multimodal large model, and dynamically generate the optimal pick-up point and route planning; The group pick-up point priority sorting and dynamic adjustment module dynamically adjusts the pick-up order and pick-up point priority by predicting the needs and behaviors of passengers in the group; The real-time group interaction and feedback module maintains real-time interaction with each passenger through mobile devices, providing dynamic updates of current pick-up points and driving routes, as well as feedback on the group pick-up sequence; The passenger group data collection module, the group behavior prediction and pattern analysis module, the multimodal large model optimization module, the group pick-up point priority sorting and dynamic adjustment module and the real-time group interaction and feedback module are interconnected.

2. The autonomous driving vehicle group pick-up optimization system based on group behavior prediction according to claim 1 is characterized in that: The passenger group data collection module includes a location data collection submodule, a demand input submodule and a behavior data integration submodule; The location data acquisition submodule obtains the current location of the passenger through GPS, Wi-Fi signal, and Bluetooth; The demand input submodule is used for passengers to input pick-up and drop-off requirements through mobile devices, including specific pick-up and drop-off points, time limits, and preferred routes; The behavior data integration submodule performs personalized analysis based on the passenger's historical behavior data, which includes riding habits, frequently used pick-up and drop-off points, and preferred driving routes.

3. The autonomous driving vehicle group pick-up optimization system based on group behavior prediction according to claim 2 is characterized in that: The specific implementation process of the position data acquisition submodule is as follows: L1: Data collection, obtaining outdoor location information of passengers through GPS and generating GPS data; building a fingerprint database by collecting Wi-Fi signal characteristics for indoor location information of passengers and generating Wi-Fi data; Through Bluetooth beacons, the beacon broadcast data packets received by the passenger's terminal device are collected to obtain location information and generate BLE data, wherein the beacon broadcast data packets at least include a MAC address and RSSI signal strength; L2: Preprocessing, preprocessing the collected GPS data, Wi-Fi data and BLE data, including noise filtering and data synchronization; L3: Data fusion, which integrates GPS data, Wi-Fi data, and BLE data through the Kalman filter algorithm; The specific implementation process of L3 is as follows: L31: Set the initial state and covariance matrix of the Kalman filter, including the process noise covariance and the observation noise covariance; L32: Predict the current state and covariance based on the previous state; L33: Update state estimates and covariances using new observations; L34: Repeat steps L32 and L33 to achieve continuous tracking and optimization of passenger positions; The specific implementation process of the demand input submodule is as follows: Passengers submit pick-up requests through mobile devices. Natural language processing technology is used to process the pick-up requests submitted by passengers. The pre-trained BERT language model is used to analyze the demand description in the pick-up request of the passenger, and the demand description is converted into a pick-up demand. The named entity recognition (NER) technology is used to identify the geographic location and time in the pick-up demand, and the geographic location and time are converted into parameters in the pick-up plan. The specific implementation process of the behavior data integration submodule is as follows: Using the historical behavior data of passengers, a hidden Markov model is used to model the passenger behavior and generate a prediction model for the passenger's future behavior; The input of the passenger future behavior prediction model training is the passenger's historical behavior data, and the output is the passenger's future behavior pattern, which includes whether the passenger is expected to show up on time, the passenger's preferred route, and the passenger's common behavior in a specific time period; The hidden Markov model formula is: P(A|θ)=∑ allstates P(A,Q|θ), Among them, A is the observed behavior sequence, θ is the model parameter, which is obtained by maximum likelihood estimation, and Q is the state of the behavior sequence.

4. The autonomous driving vehicle group pick-up optimization system based on group behavior prediction according to claim 3 is characterized in that: The group behavior prediction and pattern analysis module includes a behavior pattern analysis submodule, a trajectory prediction submodule and a group behavior prediction submodule; The behavior pattern analysis submodule identifies the behavior pattern of the passenger by analyzing the historical behavior data of the passenger, wherein the behavior pattern at least includes a preference for getting on and off the bus at certain specific times and locations; The trajectory prediction submodule predicts the passenger's future movement path based on the current location and environmental information; The group behavior prediction submodule generates the best pick-up and drop-off sequence and path for the entire group based on the behavior patterns of multiple passengers, and adjusts the plan according to the needs of different passengers.

5. The autonomous driving vehicle group pick-up optimization system based on group behavior prediction according to claim 4 is characterized in that: The specific implementation process of the behavior pattern analysis submodule is as follows: Based on the historical behavior data of passengers, the clustering algorithm K-means is used to analyze the behavior patterns of different passengers, and classification is performed based on the results of the pattern analysis; The specific implementation process of the trajectory prediction submodule is as follows: The passenger's movement trajectory is predicted using a spatiotemporal trajectory prediction model based on a long short-term memory network (LSTM), wherein the training input of the spatiotemporal trajectory prediction model is the passenger's historical trajectory, and the output is the passenger's future movement trajectory; The specific implementation process of the group behavior prediction submodule is as follows: A model combining graph convolutional network (GCN) and Transformer attention mechanism is used to globally optimize the behaviors of multiple passengers. The graph convolutional network (GCN) predicts the overall behavior pattern by processing the relationship network between group members, and Transformer is used to capture the complex dependencies of the relationship network to generate the optimal pick-up and drop-off sequence and path plan. The specific implementation process of the optimal pick-up sequence and path plan is as follows: S1: Multiple passenger service area division based on K-Means++ algorithm; S2: Determination of multiple passenger sharing points based on DBSCAN algorithm; S3: Generation of optimal pick-up and drop-off sequence paths based on maximum likelihood estimation; The specific implementation process of S1 is as follows: S11: Input a coordinate data set N containing n passenger demand points, N = {(x i ,y i )|i=1,2,3,...,n}, input the preset category number k, where the value of k is determined according to the required quantity, that is, Q is the vehicle seating capacity; S12: Randomly select a point (x p ,y p ) as the cluster center, and calculate the shortest distance between each sample and the current cluster center using D pi Indicates that the D pi Use the third-party map API interface to obtain the longitude and latitude coordinates of the passenger demand points, and then obtain the driving distance between the two passenger demand points; S13: Calculate the probability of each sample point being selected as the next cluster center point S14: Repeat steps S12 and S13 until k cluster centers are selected; S15: Calculate the distances between the k cluster centers of each data sample, and select the cluster center with the closest distance to be classified into one category; S16: Calculate the mean distance in each cluster, update the cluster center by the mean distance between the horizontal coordinate and the vertical coordinate, and repeat steps S15 and S16 until convergence; S17: Extract clustering results, where the data sample size in each cluster is no greater than the clustering result of the passenger capacity of a single vehicle, as the single n passenger demand points; The specific implementation process of S2 is: S21: Input data set N = {(x i ,y i )|i=1,2,3,...,N}, specify the radius of Eps neighborhood as r, the minimum number of samples in MinPts neighborhood as q, where MinPts and Eps are both parameters in DBSCAN clustering algorithm; S22: Initialize the number of clusters k'=0, initialize the unvisited sample set P=D, and the multiplication point set S23: Randomly select a data sample o i =(x i ,y i ), if |N Eps (o i )|≥MinPts, then o i Is the core object, o i Add Cluster C i In the collection, C i ={o i }, and the neighborhood subset N Eps (o i ) added to C i middle; S24: Update data sample P=PC i Repeat step S23 until each data sample point is visited and the final multiplication point set C is obtained. k' ={C1,C2,...,C i }; S25: For the same cluster C i A demand point in the random selection is used as the final boarding / disembarking point; The specific implementation process of S3 is as follows: The method of univariate linear regression is adopted to fit all passenger pool points through the least squares method to generate a uniform linear regression line of the optimal pick-up and drop-off sequence. Then, the linear regression line is input into the actual road network map to calculate the intersection points through the Dijkstra algorithm. Finally, the shortest path with cluster P0 as the starting point is generated, and each node is marked as P1, P2, ..., P n .

6. The autonomous driving vehicle group pick-up optimization system based on group behavior prediction according to claim 5 is characterized in that: The multi-modal large model optimization module includes a multi-data input processing submodule, a dynamic traffic analysis submodule and a personalized optimization submodule; The multi-data input processing submodule analyzes the instructions and requirements input by the passenger through natural language processing technology, and confirms the pick-up point location in combination with the image and text data; The dynamic traffic analysis submodule dynamically optimizes the pick-up point selection and driving route by analyzing real-time traffic data, weather information and road conditions; The personalized optimization submodule generates a personalized pick-up and drop-off plan for each passenger, and the personalized pick-up and drop-off plan needs to be combined with the specific needs of the passenger and the needs of other members in the group.

7. The autonomous driving vehicle group pick-up optimization system based on group behavior prediction according to claim 6, characterized in that: The specific implementation process of the multi-data input processing submodule is as follows: Using a multimodal fusion CLIP model to process the input of passenger voice, image and text, the multimodal fusion CLIP model outputs the passenger's location information, and the passenger's location information is used to fuse the real-time image data to confirm the pick-up point; The specific implementation process of the dynamic traffic analysis submodule is as follows: Through the deep Q network model and proximal operator Proximal, real-time traffic information and road conditions are analyzed to dynamically adjust the vehicle's pick-up and drop-off routes; The specific implementation process of the personalized optimization submodule is as follows: A genetic algorithm for multi-objective optimization is used to generate an optimization solution for each passenger's personalized needs, including preferred pick-up and drop-off times and specific route preferences.

8. The autonomous driving vehicle group pick-up optimization system based on group behavior prediction according to claim 7, characterized in that: The group pick-up point priority sorting and dynamic adjustment module includes a pick-up point sorting submodule and a dynamic adjustment submodule; The pick-up point sorting submodule prioritizes the pick-up points according to the needs of each passenger, wherein the needs include at least time limits and preferred locations; The dynamic adjustment submodule is used to automatically adjust the pick-up order and route when the needs or position of a passenger in the group changes, and the position change at least includes late arrival or temporary change of the pick-up point.

9. The autonomous driving vehicle group pick-up optimization system based on group behavior prediction according to claim 8, characterized in that: The specific implementation process of prioritizing the pick-up and drop-off points in the pick-up and drop-off point sorting submodule is as follows: Priority sorting uses a multi-criteria decision-making TOPSIS algorithm to generate a priority sorting of pick-up points based on passenger needs, calculate the positive ideal solution and negative ideal solution of each pick-up point, and generate a heat map based on the positive ideal solution and negative ideal solution. According to the distribution of each pick-up point in the heat map, dense distribution has a high score, and sparse distribution has a low score. The best pick-up order is determined based on the scores. The passenger's needs include time constraints and location preferences; The specific implementation process of the dynamic adjustment submodule is as follows: The pickup sequence is adjusted in real time using the reinforcement learning algorithm DQN. When passenger demand or location changes, the pickup sequence and path are recalculated based on the current environment, and the plan is adjusted based on the latest traffic and passenger data. The current environment includes passenger lateness and temporary changes to the pickup point.

10. The autonomous driving vehicle group pick-up optimization system based on group behavior prediction according to claim 9, characterized in that: The real-time group interaction and feedback module includes a real-time notification submodule and a feedback and adjustment submodule; The real-time notification submodule is used to notify each passenger of the update of the pick-up point, estimated arrival time, and vehicle location through the application APP, and allow passengers to view the pick-up order and route in the group in real time; The feedback and adjustment submodule is used for passengers to provide real-time feedback or adjust their pick-up and drop-off requirements through the APP, and dynamically adjust the group pick-up and drop-off plan based on the feedback or adjustment of the pick-up and drop-off requirements.

Images