Driving intention recognition improvement method, device and equipment based on heavy truck fuel-saving cruise control and storage medium

Abstract

The invention discloses a driving intention recognition improvement method, device and equipment based on heavy truck fuel-saving cruise control and a storage medium, relates to the technical field of vehicle fuel-saving control, and is used for solving the problems of low fuel efficiency, inaccurate interaction intention recognition and high computing resource consumption in the prior art. The method comprises the following steps: constructing a prediction model of fusion of a dynamic graph neural network and an LSTM, obtaining interaction information of surrounding vehicles, and extracting potential driving intention features; on the basis of the prediction result, balance control between fuel saving and safety is achieved through a multi-stage mode switching mechanism; an event-triggered attention distribution method is introduced, and a prediction area is segmented to improve the calculation efficiency; and in combination with real-time road network data, solving of an oil consumption optimization control strategy is completed based on a quadratic programming method. According to the method, the prediction accuracy of the vehicle driving intention can be improved, compared with an existing self-adaptive cruise control system, fuel consumption per kilometer can be reduced, and the method has good energy saving performance, stability and practical application value.

Description

Technical Field

[0001] This patent relates to the field of cruise control and fuel-saving control of commercial vehicles, and particularly to the problem of predicting the driving intention of commercial vehicles. Specifically, it relates to an improved method, device, equipment and storage medium for recognizing driving intention based on fuel-saving cruise control of heavy commercial vehicles. Background Technology

[0002] With the development of intelligent transportation systems, autonomous driving and advanced driver assistance systems are increasingly being applied in the commercial vehicle sector. Heavy-duty commercial vehicles, as the mainstay of road transportation, account for a significant proportion of enterprise operating costs due to fuel consumption. Therefore, achieving efficient, energy-saving, and safe driving control in real-world road environments has become a key research focus. Currently, mainstream fuel-saving cruise control systems (Predictive Adaptive Cruise Control, PACC) largely rely on fixed routes, static rules, or preset controller parameters, making it difficult to adapt to complex and changing traffic environments and the behavior of vehicles ahead.

[0003] Traditional methods typically rely on pre-defined vehicle following models and rule-based controllers, lacking the ability to effectively identify and predict the intentions of the vehicle ahead. Furthermore, in complex situations such as multi-vehicle interactions, incline roads, and unexpected events, existing systems often exhibit delayed responses, making it difficult to balance fuel economy and driving safety. Meanwhile, with the development of deep learning and graph neural networks, effectively combining them with time-series predictive models and applying them to vehicle control systems has become a critical issue that urgently needs to be addressed.

[0004] Therefore, it is necessary to build a new type of fuel-saving control system that can dynamically model vehicle interaction relationships, accurately predict the intentions of surrounding vehicles, and adjust control strategies in real time based on road information, thereby improving the system's adaptability, stability, and energy-saving performance.

[0005] The above content is only used to help understand the technical solution of this application and does not represent an admission that the above content is prior art. Summary of the Invention

[0006] The main objective of this application is to provide a hierarchical cruise control method, device, equipment, and storage medium based on the driving intentions of surrounding vehicles, aiming to solve the technical problem of how to take into account changes in the driving intentions of surrounding vehicles and plan a safe and economical vehicle speed to achieve efficient fuel saving.

[0007] To achieve the above objectives, this application proposes a hierarchical cruise control method based on the driving intentions of surrounding vehicles, the method comprising:

[0008] Based on the driving information of surrounding vehicles when an emergency occurs, the potential driving behavior of these vehicles is predicted, and the driving intention prediction results of surrounding vehicles are obtained.

[0009] The vehicle cruise control mode is determined based on the relative distance to the vehicle ahead, relative speed, and / or the predicted driving intentions of surrounding vehicles. Furthermore, the vehicle control mode includes an energy-saving mode, an adaptive mode, and a safety mode.

[0010] Fuel-saving control is achieved based on the vehicle cruise control mode.

[0011] In one embodiment, the predictive adaptive cruise control method based on dynamic graph-LSTM includes the following steps:

[0012] Based on the vehicle's operating status and traffic environment information, a cruise control mode is determined, which includes an economy mode, an adaptive mode, and a safety mode.

[0013] In the adaptive mode, a dynamic graph structure is constructed, and the vehicle and surrounding target vehicles are modeled as nodes in a graph neural network, with the dynamic interaction relationships between nodes as edge vectors.

[0014] The graph neural network is used to extract interaction pattern features between surrounding vehicles and generate intermediate potential driving intentions.

[0015] The intermediate latent intent representation is input into the LSTM network to complete the sequence prediction of the future motion state of surrounding vehicles and generate multi-temporal driving intent information.

[0016] When the identified preceding vehicle's intent is a specific event such as acceleration, deceleration, or lane change, the intent region is divided using an attention mutation mechanism, and computational resource allocation is optimized based on the region's confidence level.

[0017] Combining the vehicle's current power status and the identified future intentions, fuel-saving optimization control commands are generated using preset road curvature information and slope models, including desired acceleration, shift strategy and energy recovery configuration;

[0018] Based on the control commands, the vehicle performs predictive energy-saving cruise control to minimize energy consumption and ensure operational safety in changing traffic scenarios.

[0019] In one embodiment, the step of determining the cruise control mode based on the vehicle operating status and traffic environment information includes:

[0020] The road network is uniformly divided based on the complete road dataset and then combined with information signals from various vehicle chassis to form a road network matrix.

[0021] A fuel consumption model is constructed based on the road network matrix;

[0022] The fuel consumption optimization target is set based on the fuel consumption model.

[0023] In one embodiment, the step of constructing a dynamic graph structure in the adaptive mode, modeling the vehicle and surrounding target vehicles as nodes in a graph neural network, and using the dynamic interaction relationships between nodes as edge connections, includes:

[0024] Traffic event data and driving data of surrounding vehicles are obtained based on the sensor's sensing range information;

[0025] The node vectors and edge vectors are calculated based on the traffic event data and the driving data of surrounding vehicles using a dynamic graph network.

[0026] In one embodiment, the step of extracting interaction pattern features between surrounding vehicles based on a graph neural network and generating intermediate latent intentions includes:

[0027] The driving intention of vehicles within a defined area is captured based on a mutational temporal self-attention mechanism;

[0028] The parameters are updated based on the adjacency matrix update function, the backpropagation function, and the cross-entropy loss function.

[0029] Based on the driving intentions of vehicles within the aforementioned area and the updated parameters, the interaction features between vehicles are extracted.

[0030] The interaction between the vehicle and all surrounding vehicles is calculated based on the relative speed and relative distance between the vehicle and surrounding vehicles.

[0031] The intermediate potential intentions are calculated based on the above interactions.

[0032] Furthermore, to achieve the above objectives, this application also proposes a graph neural network prediction architecture that incorporates an attention mutation mechanism, the graph neural network prediction architecture incorporating the attention mutation mechanism comprising:

[0033] The sudden event detection module is used to automatically divide the prediction region and identify local abnormal changes during the training or inference process of graph neural networks by using the abrupt change characteristics of the dynamic attention distribution between nodes.

[0034] The region partitioning module is used to adaptively adjust the aggregation method of nodes in the graph structure based on the results of sudden event detection, thereby narrowing the prediction calculation range and improving computational efficiency.

[0035] The graph convolution prediction module is used to perform spatiotemporal graph convolution operations on local subgraphs and generate prediction results.

[0036] Furthermore, to achieve the above objectives, this application also proposes a graph neural network prediction device that incorporates an attention mutation mechanism. The device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. The computer program is configured to implement the steps of the graph neural network prediction method with an attention mutation mechanism as described above.

[0037] In addition, to achieve the above objectives, this application also proposes a storage medium, which is a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the steps of the graph neural network prediction method with an attention mutation mechanism as described above.

[0038] In addition, to achieve the above objectives, this application also provides a computer program product, which includes a computer program that, when executed by a processor, implements the steps of the graph neural network prediction method with an attention mutation mechanism as described above.

[0039] One or more technical solutions proposed in this application have at least the following technical effects:

[0040] By introducing an attention mutation mechanism, the mutation behavior of attention weights in graph structures can be perceived in real time, enabling dynamic partitioning and compression of the prediction region, reducing the consumption of redundant computing resources, while maintaining prediction accuracy. It is suitable for graph structure modeling scenarios with dynamic adjacency relationships and has good generalization ability and efficiency advantages. Attached Figure Description

[0041] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0042] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1 This is a flowchart of the overall method of this patent;

[0044] Figure 2 This is a diagram of the attention mechanism at an intersection;

[0045] Figure 3 This is a diagram of the LSTM model architecture;

[0046] Figure 4 This is a diagram illustrating safe mode cruise control;

[0047] Figure 5 This is a diagram of dynamic programming;

[0048] Figure 6 This is a schematic diagram of the device structure of the hardware operating environment involved in the hierarchical cruise control method based on the driving intentions of surrounding vehicles in the embodiments of this application;

[0049] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0050] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0051] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0052] Reference Figure 1 A hierarchical cruise control method, device, equipment, and storage medium based on the driving intentions of surrounding vehicles. The working process of this invention is as follows:

[0053] 1. Data Preparation and Processing

[0054] Step 1: Acquire Vehicle Interaction Data. Considering the relatively small differences in driving behavior and dynamic characteristics between passenger cars and heavy commercial vehicles, data is collected using passenger cars. Based on speed range, vehicle dynamic data is divided into three categories: low speed (30-50 km / h), medium speed (50-70 km / h), and high speed (70-90 km / h). High-speed data is collected using high-fidelity driving simulation software that supports centimeter-level precision modeling and data import; the remaining data comes from passenger car sensors.

[0055] Step Two: Collect and process road information data. First, acquire road information of heavy commercial vehicles traveling at high speeds (70 km / h) and construct a road dataset. Due to the high noise level in the raw data, data cleaning and filtering techniques are used for preprocessing to obtain a streamlined, high-quality road dataset, thereby improving the accuracy and stability of model training and control strategy formulation.

[0056] 2. Intent Prediction Network Modeling

[0057] Step 1: Dynamic Graph Model Construction. To construct this model, the raw data such as position and velocity need to be converted into a graph, denoted as G(V,E), where V represents the node vector and E represents the edge vector. The formula for the eigenvector of a node is shown below. Theoretically, there exists an adjacency matrix between every pair of vehicles within the region. However, considering the computational limitations of the onboard controller, the adjacency matrix of the directed graph is only established within the prediction region. Where x...n It is the feature vector of node n.

[0058]

[0059] Step 2: Modeling the Intent Interaction Network.

[0060] (1) such as Figure 2 As shown, a mutation-type temporal self-attention mechanism is established to capture the movement of vehicles within a defined region, thereby simplifying the training process. The expression is as follows:

[0061]

[0062] Among them, 𝛼 𝑖𝑗 E represents the attention weight from node j to node i; st,i and E st,j It is a spatiotemporal embedding that encodes the features of nodes i and j respectively; α T These are feature weights; [W] e E st,i ||W e E st,j ] represents the feature concatenation operation; f() represents the nonlinear transformation function; N(i) represents the set of neighbors of node i.

[0063] (2) Parameter update: The parameters are updated using the adjacency matrix update function A', the backpropagation function L, and the cross-entropy loss function b, as shown in the following expression:

[0064]

[0065] Where λ is the weighting factor and η is the learning rate.

[0066] (3) Interaction Feature Extraction: The interaction features between vehicles are extracted using a feature extraction function, the specific form of which is shown in the following formula:

[0067]

[0068] Among them, W i and W f It is the weight factor matrix, F i Let be the aggregate feature of node i, and be the activation function.

[0069] (4) Integration of vehicle dynamics and interaction: There is an interaction between the vehicle and all surrounding vehicles, and the strength of this interaction depends on the relative speed w. ij and relative distance d Iref Its expression is as follows:

[0070]

[0071] Among them, 𝛼 𝑖𝑗 d is the interaction strength weight from node i to node j. Lar It is the weighted relative distance, w La and w Lo These are the weights for horizontal and vertical distances, respectively.

[0072] (5) Interaction model construction: Based on the above w ij and I ij Establish interaction model F interact,i The interaction features between the vehicle and all vehicles within the prediction area are extracted, and their expression is shown below:

[0073]

[0074] (6) Vehicle driving intention recognition: Refer to Figure 3 Based on the aforementioned interaction characteristics, a Long Short-Term Memory (LSTM) network is used as the network model for recognizing the target vehicle's driving intention. Figure 3 As shown, LSTM consists of gate and memory cells, C t C represents the information stored in the cell at time t. t-1 This represents the storage unit information at time t-1. This represents candidate memory cell information, indicating new memory information at the current time. h t h represents the hidden layer output information at time t. t-1 This represents the hidden layer output information at time t-1. t Represents the forget gate, used to determine C t-1 How much of the information in it has been forgotten? t This represents the input gate, used to determine the importance of the input information at time t. t The output gate is represented by σ, used to obtain the importance of the output information at time t. σ and tanh are activation functions used to introduce nonlinear characteristics; σ is the sigmoid function, which maps the input to the range of 0 to 1, and tanh is the hyperbolic tangent function, which maps the input to the range of -1 to 1. t Let y be the input vector at time t, representing the temporal feature vector obtained after temporal encoding of the chassis information of the preceding vehicle. t Let be the output vector at time t, representing the predicted probability distribution of the preceding vehicle's lane change intention and speed change intention. The memory cell output expression of the LSTM unit is as follows:

[0075]

[0076] 3. Layered fuel-saving cruise control modeling

[0077] (1) Establish a map processing module: Divide the complete road dataset into 200m intervals according to the slope, and then combine it with information signals from various vehicle chassis to form a new road network matrix.

[0078] (2) Fuel Consumption Model Construction: This invention proposes a fuel consumption estimation model based on stage-based path, the core of which lies in calculating the fuel consumption of each stage through vehicle operating parameters, for path prediction and energy-saving strategy optimization. Wherein, 𝑄 𝑡（𝑘） This indicates the main fuel consumption indicator for the forecast phase, with the forecast path divided at equal intervals, each interval set at 20 meters; （𝑘） R is the vehicle's speed during stage X. 𝑡（1） This represents the fuel consumption rate of the first segment of the predicted path, obtained by dividing the path into equal intervals of 200 meters. 𝑒（𝑘） For engine speed, 𝑇 𝑒（𝑘） The engine torque is given by , and the radius of action is given by , which is equal to the tire radius divided by the transmission efficiency and the gearbox ratio. Vehicle force factors include gravitational drag, gradient effect, air resistance, and acceleration drag, where is the vehicle mass, is the acceleration due to gravity, is the rolling resistance coefficient, and is the total resistance coefficient. （𝑘） Where is the stage slope angle, is the acceleration correction factor, and is the acceleration angle. （𝑘） For acceleration, 𝐶 𝐷 Let be the drag coefficient, be the frontal area, and be the air density. The model's fuel consumption calculation formula is as follows:

[0079]

[0080] (3) Fuel Consumption Optimization Target Construction: Optimization control targets are set based on fuel consumption, deviation between driving speed and expected speed, speed changes, and total driving time. Furthermore, a penalty for deviation of the final speed from the set speed is introduced as the final penalty target. The constructed optimization model is shown in the following equation:

[0081]

[0082] Where, 𝐽 represents the total cost function of the problem, 𝜆1 represents the weight coefficient of fuel economy in the dynamic planning (DP) algorithm, and 𝜆2 represents the weight coefficient of the terminal penalty term in the dynamic planning algorithm, specifically the vehicle speed error penalty coefficient.

[0083] (4) Construction of a layered cruise system model: Three cruise modes are established based on fuel consumption and fuel consumption optimization objectives. These are:

[0084] Safety Mode: As the highest-priority key module in the vehicle control framework, safety mode is primarily used for collision avoidance. This paper proposes a collision avoidance system that integrates the driving intentions of surrounding vehicles, determining whether to execute braking actions by setting a risk threshold. Figure 4 As shown, the safety mode is also triggered when a vehicle in an adjacent lane intends to merge into this lane. Specifically, [the following is unclear and likely refers to a specific feature or function]. B Indicates the applied braking acceleration, 𝑎 𝑚𝑎𝑥 T is the maximum permissible braking acceleration of the vehicle. e T represents the current environmental risk value. rep t is the risk threshold. e T represents the minimum braking time. t The braking judgment formula for the safety mode is as follows, given the time step of the intent prediction module:

[0085]

[0086] Fuel-saving mode: The energy-saving mode is based on an extension of the map processing module, which plans a series of optimal energy-saving speeds for future road segments using a linear interpolation method. The planning process follows preset constraints. This module outputs the energy-saving speed corresponding to the current vehicle position. To ensure computational efficiency, the module uses a dynamic programming (DP) algorithm for optimization. The solution process of the DP algorithm is as follows: Figure 5 As shown. The system state variable is x(k) = [ω e (k) T e (k) v(k) a(k)] T The control variable is the driving speed u(k) = φ(k). This paper solves for the final driving speed φ(0) - φ(10) using a recursive formula, ultimately forming the planned economic speed sequence φ(0) - φ(10). The vehicle speed constraint is φ, the rotational speed constraint is ω, the torque constraint is T, and the acceleration constraint is α. The optimization objective function expression of the dynamic programming algorithm is:

[0087]

[0088]

[0089] Where 𝐽 is the total cost function, 𝑓 is the single-stage cost function, 𝑔 is the immediate cost, ℎ is the state transition function, 𝑘 is the discrete time step, and 𝑁 is the planning length.

[0090] Adaptive Mode: The Generality Mode is based on Predictive Adaptive Cruise Control (PACC) and is primarily applied to energy-efficient driving optimization in medium-to-high traffic density scenarios. This mode comprehensively considers road gradient and driving safety factors, using the vehicle speed sequence generated by the energy-efficient mode as the reference speed input for the vehicle, guiding the PACC speed planning module to perform further optimization. Specifically, this reference speed, combined with distance information, is input to the secondary speed planning process based on Model Predictive Control (MPC), forming the algorithm framework for PACC speed planning. Within this framework, the basic structure of the algorithm model is defined as follows: where x(k) represents the system state variable, u(k) represents the control input variable (in this study, the desired acceleration), w(k) represents the external disturbance of the system, i.e., the acceleration of the vehicle in front, and y(k) represents the system output variable. The matrix representation of the system state-space model is as follows:

[0091]

[0092]

[0093]

[0094]

[0095]

[0096]

[0097] This embodiment determines the planning time by acquiring real-time information on the driving of surrounding vehicles, and accurately predicts the potential driving intentions of surrounding vehicles based on their relative distance, relative speed, and interaction characteristics. Simultaneously, it combines layered cruise control with maintaining a safe distance from surrounding vehicles to ensure driving safety and optimize fuel economy, effectively improving driving comfort and the vehicle's intelligent control level.

[0098] It should be noted that the above examples are only for understanding this application and do not constitute a limitation on the layered cruise control method based on the driving intentions of surrounding vehicles in this application. Any simple modifications based on this technical concept are within the protection scope of this application.

[0099] The hierarchical cruise control device based on the driving intentions of surrounding vehicles provided in this application, employing the hierarchical cruise control method based on the driving intentions of surrounding vehicles in the above embodiments, can solve the technical problem of how to consider changes in vehicles in the lane ahead and plan a safe and economical speed. Compared with the prior art, the beneficial effects of the hierarchical cruise control device based on the driving intentions of surrounding vehicles provided in this application are the same as those of the hierarchical cruise control method based on the driving intentions of surrounding vehicles provided in the above embodiments, and other technical features in the hierarchical cruise control device based on the driving intentions of surrounding vehicles are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0100] This application provides a hierarchical cruise control device based on the driving intention of surrounding vehicles. The hierarchical cruise control device based on the driving intention of surrounding vehicles includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the hierarchical cruise control method based on the driving intention of surrounding vehicles in the above embodiment 1.

[0101] The following is for reference. Figure 6 This document illustrates a structural schematic diagram of a hierarchical cruise control device based on the driving intentions of surrounding vehicles, suitable for implementing embodiments of this application. The hierarchical cruise control device based on the driving intentions of surrounding vehicles in embodiments of this application may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital radio receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. Figure 6 The layered cruise control device shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0102] like Figure 6As shown, the hierarchical cruise control device based on the driving intentions of surrounding vehicles may include a processing unit 1001 (e.g., a central processing unit, a graphics processor, etc.), which can perform various appropriate actions and processes according to a program stored in ROM (Read Only Memory) 1002 or a program loaded from storage device 1003 into RAM (Random Access Memory) 1004. RAM 1004 also stores various programs and data required for the operation of the hierarchical cruise control device based on the driving intentions of surrounding vehicles. The processing unit 1001, ROM 1002, and RAM 1004 are interconnected via bus 1005. Input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to I / O interface 1006: input devices 1007 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 1008 including, for example, LCDs (Liquid Crystal Displays), speakers, vibrators, etc.; storage devices 1003 including, for example, magnetic tapes, hard disks, etc.; and communication devices 1009. Communication device 1009 allows the hierarchical cruise control device based on the driving intentions of surrounding vehicles to communicate wirelessly or wiredly with other devices to exchange data. Although a hierarchical cruise control device based on the driving intentions of surrounding vehicles with various systems is shown in the figure, it should be understood that it is not required to implement or possess all the systems shown. More or fewer systems may be implemented alternatively.

[0103] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from ROM 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0104] The hierarchical cruise control device based on the driving intentions of surrounding vehicles provided in this application, employing the hierarchical cruise control method based on the driving intentions of surrounding vehicles in the above embodiments, can solve the technical problem of how to consider changes in vehicles in the lane ahead and plan a safe and economical speed. Compared with the prior art, the beneficial effects of the hierarchical cruise control device based on the driving intentions of surrounding vehicles provided in this application are the same as those of the hierarchical cruise control method based on the driving intentions of surrounding vehicles provided in the above embodiments, and other technical features in this hierarchical cruise control device based on the driving intentions of surrounding vehicles are the same as those disclosed in the previous embodiment method, and will not be repeated here.

[0105] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0106] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0107] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, the computer-readable program instructions being used to execute the hierarchical cruise control method based on the driving intentions of surrounding vehicles in the above embodiments.

[0108] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, RAM (Random Access Memory), ROM (Read Only Memory), EPROM (Erasable Programmable Read Only Memory), or flash memory, optical fiber, CD-ROM (CD-Read Only Memory), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0109] The aforementioned computer-readable storage medium may be included in a layered cruise control device based on the driving intentions of surrounding vehicles; or it may exist independently and not be installed in a layered cruise control device based on the driving intentions of surrounding vehicles.

[0110] The aforementioned computer-readable storage medium carries one or more programs that, when executed by a hierarchical cruise control device based on the driving intentions of surrounding vehicles, cause the hierarchical cruise control device based on the driving intentions of surrounding vehicles to: predict the potential lane-changing behavior of the preceding vehicle based on the chassis information of the preceding vehicle, and obtain a prediction result of the preceding vehicle's driving intention; determine a vehicle control mode based on the relative distance to the preceding vehicle and / or the prediction result of the preceding vehicle's driving intention, wherein the vehicle control mode includes a predictive cruise control mode and a following control mode; and complete predictive adaptive cruise control based on the vehicle control mode.

[0111] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including LAN (Local Area Network) or WAN (Wide Area Network)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0112] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0113] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0114] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the above-described hierarchical cruise control method based on the driving intentions of surrounding vehicles. This solves the technical problem of how to consider changes in vehicle conditions in the lane ahead and plan for a safe and economical vehicle speed. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the hierarchical cruise control method based on the driving intentions of surrounding vehicles provided in the above embodiments, and will not be repeated here.

[0115] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the hierarchical cruise control method based on the driving intentions of surrounding vehicles as described above.

[0116] The computer program product provided in this application can solve the technical problem of how to plan a safe and economical vehicle speed by taking into account changes in vehicles in the lane ahead. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as those of the hierarchical cruise control method based on the driving intentions of surrounding vehicles provided in the above embodiments, and will not be repeated here.

[0117] The above description is only a part of the embodiments of this application and does not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. An improved method, apparatus, device, and storage medium for recognizing driving intentions based on fuel-saving cruise control for heavy-duty trucks, characterized in that, include: S1, Data Acquisition Module, used to acquire real-time operating status information of the target vehicle and surrounding vehicles, as well as road environment information; S2, Dynamic Graph Construction Module, is used to model vehicles as graph nodes, vehicle interactions as graph edges, and dynamically generate vehicle interaction graphs. S3, Long Short-Term Memory (LSTM) prediction module, connected to the dynamic graph construction module, used for temporal intent prediction based on interaction features extracted by graph neural network; S4, a three-mode switching control module, is used to switch between energy-saving mode, adaptive mode and safety mode based on the prediction results; S5, Road Condition Optimization Module, is used to perform path fitting and segmentation by combining real-time road network information; S6, the fuel consumption optimization module, is used to solve the cruise control output based on the prediction results and the fitted road conditions, so as to realize the fuel-saving control strategy.

2. The control system according to claim 1, characterized in that, The S2 module uses a dynamic graph neural network (DGNN) to model the interaction relationships between nodes and supports real-time updates of the adjacency matrix to adapt to changes in the traffic environment.

3. The control system according to claim 1, characterized in that, The S3 module is used to perform temporal prediction of vehicle intent based on the extracted graph embedding features, with a prediction accuracy of not less than 94%.

4. The method according to claim 3, characterized in that, It further includes an event-triggered attention allocation module, which is used to trigger the reallocation of attention weights when a specific event occurs, thereby dynamically dividing the prediction region and reducing invalid computation.

5. The control system according to claim 1, characterized in that, The S4 module includes: S41, Energy Saving Mode, is used to employ the optimal fuel consumption control strategy when operating conditions are stable; S42, Adaptive mode, used to adjust control parameters according to the intentions of surrounding vehicles under moderate traffic complexity; S43, Safety Mode, is used to prioritize vehicle safety when there is a potential collision risk.

6. The control system according to claim 1, characterized in that, The S5 module fits and models road slope, lane width, speed limit information, etc. based on high-precision map data, and inputs the results into the fuel consumption optimization module.

7. The control system according to claim 1, characterized in that, The S6 module is based on a quadratic programming algorithm to minimize fuel consumption per unit distance while satisfying vehicle dynamic constraints and safety boundaries.

Images