A vehicle speed prediction method and system for urban scenes

Abstract

The present invention relates to a vehicle speed prediction method and system for urban scenarios. The method uses a personalized federated learning method to achieve personalized aggregation on the server, capture client-specific information, and perform personalized vehicle speed prediction through a sequence-to-sequence long short-term memory network model based on multi-head attention. The personalized federated learning method includes: the server sends a global model to the client to initialize the local model; the client performs iterative training and uploads the model to the server; the server aggregates the local model to update the global model, and decides whether to calculate the personalized aggregation weight based on the number of iterations; if calculation is required, the personalized aggregation weight is obtained based on the weighted mean square error, and then a client-customized aggregation model is generated and sent. The network model adopts an encoder-decoder architecture, including two layers of long short-term memory network and a multi-head attention layer, and finally a fully connected layer. The present invention can achieve personalized vehicle speed prediction while protecting data privacy.

Description

Technical Field

[0001] The present invention belongs to the technical field of vehicle speed prediction and relates to a vehicle speed prediction method and system for use in urban scenarios. Background Art

[0002] Vehicle speed prediction plays an important role in modern intelligent transportation systems and is a key technology for improving road safety, traffic efficiency, and vehicle energy efficiency. It is widely used in applications such as route planning, energy management for hybrid electric vehicles, and eco-adaptive cruise control. However, accurately predicting the speed of individual vehicles is challenging because vehicle speed is affected by traffic conditions, vehicle type, road conditions, and driver behavior. Traditional methods for vehicle speed prediction include Markov chains and Bayesian networks. Compared to traditional prediction methods, recurrent neural networks leveraging vehicle-to-vehicle and vehicle-to-infrastructure communication have been shown to be more effective in vehicle speed prediction. However, previous studies have failed to account for significant differences in driver driving styles and vehicle types, which significantly impact speed prediction for target vehicles and thus prevent personalized prediction. Collaborative training of driving data collected from multiple vehicles is crucial to improving prediction accuracy, as a single vehicle does not have sufficient data to train a reliable model. However, traditional distributed machine learning techniques require the aggregation of all vehicles' private driving data on a central server (e.g., a cloud server), which poses a potential risk of private data leakage.

[0003] Federated learning, a collaborative distributed machine learning paradigm, coordinates learning tasks from a central server and is solved collaboratively by a network of participating devices (called clients). Each client maintains its own private training dataset, which is not uploaded to the server. Instead, locally trained models are uploaded to the server for aggregation into a global model, effectively reducing the privacy risks of the system. However, when applied to vehicle speed prediction, traditional federated learning methods fail to achieve personalized vehicle speed predictions using a single global model.

[0004] Personalized federated learning methods can address statistical heterogeneity in federated learning and provide personalized solutions. Unlike traditional federated learning, which seeks a single high-quality global model, personalized federated learning methods are proposed to prioritize training each client's local model.

[0005] However, existing personalized federated learning methods almost always require additional steps on the client side to achieve personalization, which often increases client-side computational or communication overhead. Given the high real-time requirements of driving scenarios and the limited computational and communication resources of vehicles, implementing the personalization step on a powerful server is undoubtedly more appropriate. Therefore, a federated learning method that implements personalization on the server is urgently needed for personalized vehicle speed prediction. Summary of the Invention

[0006] In view of this, the object of the present invention is to provide a vehicle speed prediction method and system for urban scenarios, which captures client-specific information by implementing a personalized aggregation step on the server and performs personalized prediction of vehicle speed through a new prediction model.

[0007] In order to achieve the above object, the present invention provides the following technical solutions:

[0008] A vehicle speed prediction method for urban scenarios is proposed. In this method, a personalized federated learning method is used to implement a personalized aggregation step on the server to capture client-specific information. The acquired data information is used to perform personalized prediction of vehicle speed through a sequence-to-sequence long short-term memory network model based on multi-head attention.

[0009] Furthermore, the personalized federated learning method includes the following steps:

[0010] S1. The server sends the initialized global model to all clients and initializes the local model of each client. The client is numbered i and the number of iterations of the current personalized federated learning method is t.

[0011] S2: Client i performs the tth iteration, trains the local model by gradient descent, and uploads the trained local model to the server;

[0012] S3. The server performs global aggregation on the uploaded local model to obtain a global model.

[0013] S4. The server determines whether it is necessary to calculate the personalized aggregation weight based on whether the current number of iterations t is greater than the set threshold r. If not, the server sends the global model to the selected client as its initial local model and returns to step S2. If it is greater, the server proceeds to step S5.

[0014] S5. The server calculates the weighted mean square error between the client's local model and the global model parameters and normalizes it to obtain the personalized aggregation weight for this iteration.

[0015] S6. The server obtains the client-customized aggregated model based on the personalized aggregation weight by aggregating the client's local model and the highest p-layer network of the global model at the element level, while the remaining lower-layer networks adopt the corresponding parts of the global model;

[0016] S7. The server sends the client-customized aggregation model to the selected client as its initialization local model;

[0017] S8. Repeat steps S2 to S7 for T rounds of iterations, t∈[0, T], until the local model training of each client converges.

[0018] Furthermore, in step S1, the global objective function of federated learning is defined as:

[0019]

[0020] in, and and is the local objective function of the i-th client; D i is the dataset of the i-th client; |D i | is the dataset size of the i-th client; Initialize the local model on the i-th client.

[0021] Furthermore, in step S2, for the tth round of iterative training, the server randomly selects a subset S of N clients t Perform training and send the global model or aggregate model obtained in the previous iteration to the set S t The client in

[0022] Client i performs the tth iteration and trains the local model by gradient descent as follows:

[0023]

[0024] Where η is the learning rate for local model training.

[0025] Furthermore, in step S3, the server performs global aggregation on the uploaded local model to obtain a global model, specifically in the following manner:

[0026]

[0027] Further, in step S5, the client subset S is calculated t All local models and p higher layers of the global model: and Θ t，p The weighted mean square error of the parameters is used to obtain the model parameter difference measure M t，p, which reflects the degree of difference in model parameters between different clients, and this difference is weighted based on the number of local data samples, as follows:

[0028]

[0029] M t，p Each layer After normalization, the aggregation weight W is obtained t，p , making The specific method is as follows:

[0030]

[0031] Furthermore, in step S6, it includes:

[0032] The server is based on the personalized aggregation weight W t，p , by aggregating client-side local models at the element level and the global model Θ t The highest p-layer network is used, while the remaining lower-layer networks adopt the corresponding parts of the global model to obtain the client-customized aggregation model, which is as follows:

[0033]

[0034] Where ⊙ is the Hadamard product, which means the elements at the same position of two matrices are multiplied; L(Θ i )yes The number of layers, and The lower layer of The elements are 0; the aggregation weight W t，p Same shape as the remaining p higher layers.

[0035] Furthermore, the multi-head attention-based sequence-to-sequence long short-term memory network model specifically includes: predicting the vehicle speed in the future based on a period of historical input feature data; the model adopts an encoder-decoder architecture; in addition, the model adopts a long short-term memory network and a multi-head attention mechanism; specifically, the model first processes the input sequence through a two-layer long short-term memory network as an encoder to extract a high-dimensional representation of the time series data; then, through a customized multi-head attention layer, the sequence dependencies are captured in multiple subspaces in parallel to enhance the understanding of the complex patterns of the time series; next, the output of the multi-head attention module is sent to a two-layer long short-term memory network as a decoder to gradually construct a prediction of the future vehicle speed; finally, the output of the decoder is mapped to the predicted vehicle speed through a fully connected layer.

[0036] The present invention also provides a vehicle speed prediction system for urban scenes.

[0037] The beneficial effects of the present invention are:

[0038] The present invention provides a vehicle speed prediction method and system for urban scenarios, including a personalized federated learning method based on personalized aggregation weights, which performs personalized aggregation on a server to capture client-specific information. The method sends a customized aggregation model to the client to achieve personalization without incurring additional computational and communication overhead for the client. In addition, the method also includes a sequence-to-sequence long short-term memory network model based on multi-head attention and its input feature construction method, which are used to replace traditional vehicle speed prediction models. By providing a new personalized federated learning method and a new vehicle speed prediction model, the present invention can achieve personalized vehicle speed prediction while protecting driver data privacy.

[0039] Other advantages, objects, and features of the present invention will be described in part in the following description and, in part, will be apparent to those skilled in the art upon examination of the following description or may be learned from practice of the present invention. The objects and other advantages of the present invention may be realized and obtained through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be described in detail below with reference to the accompanying drawings, in which:

[0041] Figure 1 A flowchart of a personalized federated learning method based on personalized aggregate weights provided in Example 1 of the present invention;

[0042] Figure 2 A system framework diagram of a personalized federated learning method based on personalized aggregate weights provided in Example 1 of the present invention;

[0043] Figure 3 A scenario diagram for predicting vehicle speed through vehicle-to-vehicle and vehicle-to-infrastructure communications, provided in the second embodiment of the present invention;

[0044] Figure 4 A schematic diagram of the structure of a sequence-to-sequence long short-term memory network model based on multi-head attention provided in Example 2 of the present invention. DETAILED DESCRIPTION

[0045] The technical solution of the present invention is described in detail below with reference to the accompanying drawings.

[0046] The technical solution of the present invention includes a personalized federated learning method based on personalized aggregation weights according to the first aspect, the method comprising the following steps:

[0047] Step 1: The server sends the initialized global model to all clients and initializes the local model of each client; where i is the client number and t is the number of iterations of the current personalized federated learning method;

[0048] Step 2: Client i performs the tth iteration, trains the local model by gradient descent, and uploads the trained local model to the server;

[0049] Step 3: The server performs global aggregation on the uploaded local model to obtain a global model;

[0050] Step 4: The server determines whether it is necessary to calculate the personalized aggregation weight based on whether the current iteration number t is greater than the set threshold r. If not, the server sends the global model to the selected client as its initial local model and returns to step 2. If it is greater, it proceeds to step 5.

[0051] Step 5: The server calculates the weighted mean square error between the client's local model and the global model parameters and normalizes it to obtain the personalized aggregation weight for this iteration;

[0052] Step 6: Based on the personalized aggregation weight, the server aggregates the client's local model and the highest p-layer network of the global model at the element level, while the remaining lower-layer networks adopt the corresponding parts of the global model to obtain the client-customized aggregate model;

[0053] Step 7: The server sends the client-customized aggregation model to the selected client as its initial local model;

[0054] Step 8. Repeat steps 2 to 7 for T rounds of iteration until the local model training of each client converges; where t∈[0, T].

[0055] The second aspect of the present invention provides a sequence-to-sequence long short-term memory network model based on multi-head attention, wherein the input features and model structure of the vehicle speed prediction model include:

[0056] For a discrete time setting with an interval of 1s, the vehicle speed prediction model can predict the target vehicle's speed scalar within a certain range in the future through some input features. Through vehicle-to-vehicle and vehicle-to-infrastructure communication, the input features mainly come from the target vehicle, the preceding vehicle, and the traffic light, which are the target vehicle speed v T , speed of the preceding vehicle v P 、Is there a preceding vehicle? P 、Is there a traffic light? TL , the relative distance d between the target vehicle and the preceding vehicle P , the relative distance d between the target vehicle and the stop line of the traffic light TL, the current state of the traffic light signal s TL , the future state of the traffic light signal TL，k+1 ,...,s TL，k+H , target vehicle throttle, target vehicle brake, target vehicle steering angle, steer; where the preceding vehicle refers to the nearest preceding vehicle in the same lane, H is the predicted time range, and k represents the current discrete time scale;

[0057] The multi-head attention-based sequence-to-sequence long short-term memory network model predicts vehicle speed in the future based on a period of historical input feature data, that is, the input features are organized into the input sequence of the model in chronological order, and the future speed of the target vehicle is used as the output sequence of the model; the model adopts an encoder-decoder architecture, which has been proven to be effective in various sequence modeling tasks; in addition, the model adopts a long short-term memory network and a multi-head attention mechanism, the former is good at capturing temporal dependencies, and the latter can capture complex temporal dynamics and long-distance dependencies in the sequence; specifically, the model first processes the input sequence through a two-layer long short-term memory network as an encoder to extract a high-dimensional representation of the time series data; then, through a customized multi-head attention layer, the sequence dependencies are captured in multiple subspaces in parallel to enhance the understanding of the complex patterns of the time series; next, the output of the multi-head attention module is sent to a two-layer long short-term memory network as a decoder to gradually construct a prediction of the future vehicle speed; finally, the output of the decoder is mapped to the predicted vehicle speed through a fully connected layer.

[0058] Example 1:

[0059] Reference Figure 1 , which is a flow chart of a personalized federated learning method based on personalized aggregation weights provided in the first embodiment of the present invention;

[0060] The personalized federated learning method based on personalized aggregation weights provided in the first embodiment of the present invention is applied to vehicle speed prediction of intelligent connected vehicles, and includes the following steps:

[0061] Step 1: The server sends the initialized global model to all clients and initializes the local model of each client; where i is the client number and t is the number of iterations of the current personalized federated learning method;

[0062] Specifically, step 1 includes:

[0063] Step 1.1: Number the N intelligent connected vehicles (i.e., clients) participating in federated learning from 1 to N; record the dataset of the i-th client as D i , the size of the dataset is denoted as |D i|, and these data sets are not independent and identically distributed; the local model on the i-th client is recorded as Θ i ; The number of layers of the local model is recorded as L(Θ i ); the initialized local model on the i-th client is recorded as Let the global model be Θ and the personalized aggregation weight be W p , where p means that personalized aggregation is only performed on the last p layers of the local model; the loss function is recorded as The local objective function of the i-th client is recorded as

[0064] Step 1.2: Define the global objective function of federated learning as:

[0065]

[0066] in, and and is the local objective function of the i-th client.

[0067] Step 1.3, the server sends the initialized global model Θ 0 Initialize the client's local model for all clients.

[0068] Step 2: Client i performs the tth iteration, trains the local model by gradient descent, and uploads the trained local model to the server;

[0069] Specifically, step 2 includes:

[0070] Step 2.1: For the tth round of iterative training, the server randomly selects a subset S of N clients t Perform training and send the global model or aggregate model obtained in the previous iteration to the set S t The client in

[0071] Step 2.2: Client i performs the tth iteration and trains the local model by gradient descent. The specific method is as follows:

[0072]

[0073] Where η is the learning rate of local model training;

[0074] Step 3: The server performs global aggregation on the uploaded local model to obtain the global model. The specific method is as follows:

[0075]

[0076] Step 4: The server determines whether it is necessary to calculate the personalized aggregation weight based on whether the current iteration number t is greater than the set threshold r. If not, the server sends the global model to the selected client as its initial local model and returns to step 2. If it is greater, it proceeds to step 5.

[0077] Step 5: The server calculates the weighted mean square error between the client's local model and the global model parameters and normalizes it to obtain the personalized aggregation weight for this iteration;

[0078] Specifically, step 5 includes:

[0079] Step 5.1: Calculate the client subset S t All local models and p higher layers of the global model: and Θ t，p The weighted mean square error of the parameters is used to obtain the model parameter difference measure M t，p , which reflects the degree of difference in model parameters between different clients, and this difference is weighted based on the number of local data samples, as follows:

[0080]

[0081] Step 5.2, M t，p Each layer After normalization, the aggregation weight M is obtained t，p , making The specific method is as follows:

[0082]

[0083] Step 6: Based on the personalized aggregation weights, the server aggregates the client's local model and the highest p-layer network of the global model at the element level, while the remaining lower-layer networks use the corresponding parts of the global model to obtain the client-customized aggregate model. The specific method is as follows:

[0084]

[0085] Where ⊙ is the Hadamard product, which means the elements at the same position of two matrices are multiplied; L(Θ i )yes The number of layers, and The lower layer of The elements are 0; the aggregation weight W t，p Same shape as the remaining p higher layers.

[0086] Step 7: The server sends the client-customized aggregation model Sent to the selected client as its initial local model;

[0087] Step 8. Repeat steps 2 to 7 for T rounds of iteration until the local model training of each client converges; where t∈[0, T].

[0088] Reference Figure 2 , which is a system framework diagram of the personalized federated learning method based on personalized aggregation weights provided in the first embodiment of the present invention; specifically, during the t-th round of iterative training:

[0089] (1) First, the server randomly selects some intelligent connected cars (i.e., clients) for personalized aggregation, and then the clients use their trained local models Upload to the server;

[0090] (2) The server then averages and aggregates these client local models to obtain a global model Θ t ; Then, the server calculates the weight W based on the personalized aggregation t , aggregate the client local model with the global model at the element level to obtain a customized aggregate model

[0091] (3) Finally, the server distributes these customized aggregation models to each client for local training, thereby facilitating personalized vehicle speed prediction for intelligent connected vehicles.

[0092] Example 2:

[0093] Reference Figure 3 , which is a scene diagram of vehicle speed prediction through vehicle-to-vehicle and vehicle-to-infrastructure communication provided by the second embodiment of the present invention. Specifically, through vehicle-to-vehicle and vehicle-to-infrastructure communication, the input features mainly come from the target vehicle, the preceding vehicle and the traffic light, which are: target vehicle speed v T , speed of the preceding vehicle v P 、Is there a preceding vehicle? P 、Is there a traffic light? TL , the relative distance d between the target vehicle and the preceding vehicle P , the relative distance d between the target vehicle and the stop line of the traffic light TL , the current state of the traffic light signal s TL , the future state of the traffic light signal TL，k+1 ,...,s TL，k+H , target vehicle throttle amplitude throttle, target vehicle brake amplitude brake, target vehicle steering wheel steering angle steer; among them, the front vehicle refers to the nearest previous vehicle in the same lane, H is the predicted time range, and k represents the current discrete time scale.

[0094] Reference Figure 4, which is a schematic diagram of the structure of the sequence-to-sequence long short-term memory network model based on multi-head attention provided by the second embodiment of the present invention. Specifically, the model predicts the vehicle speed in the future based on a period of historical input feature data, that is, the input features are organized into the input sequence x of the model in chronological order. k-M+1 ,...,x k And the future speed of the target vehicle is used as the output sequence v of the model t，k+1 ,...,v t，k+H , where k represents the current discrete time scale, M represents the past time range, and H represents the predicted time range; the model adopts an encoder-decoder architecture, which has been proven effective in various sequence modeling tasks; in addition, the model adopts a long short-term memory network and a multi-head attention mechanism, the former is good at capturing temporal dependencies, and the latter can capture complex temporal dynamics and long-distance dependencies in the sequence; specifically, the model first processes the input sequence through a two-layer long short-term memory network (LSTM) as an encoder to extract a high-dimensional representation of the time series data; then, through a customized multi-head attention (MHA) layer, the sequence dependencies are captured in multiple subspaces in parallel, enhancing the understanding of the complex patterns of the time series; next, the output of the multi-head attention layer is fed into a two-layer long short-term memory network as a decoder to gradually build a prediction of the future vehicle speed; finally, a fully connected (FC) layer is used to map the output of the decoder to the predicted vehicle speed.

[0095] In summary, the present invention provides a method and system for predicting vehicle speed in urban scenarios, including a personalized federated learning method based on personalized aggregation weights, which performs personalized aggregation on the server to capture client-specific information. The method sends a customized aggregation model to the client to achieve personalization without incurring additional computational and communication overhead for the client. In addition, it also includes a sequence-to-sequence long short-term memory network model based on multi-head attention and its input feature construction method, which is used to replace the traditional vehicle speed prediction model. By providing a new personalized federated learning method and a new vehicle speed prediction model, the present invention can achieve personalized vehicle speed prediction while protecting driver data privacy.

[0096] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not limiting. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention can be modified without departing from the purpose and scope of the technical solutions, which should all be included in the scope of the claims of the present invention.

Claims

1. A vehicle speed prediction method for urban scenarios, characterized by: In this method, a personalized federated learning method is used to implement a personalized aggregation step on the server to capture client information. The acquired data information is used to make personalized predictions of vehicle speed through a sequence-to-sequence long short-term memory network model based on multi-head attention. The personalized federated learning method includes the following steps: S1. The server sends the initialized global model to all clients and initializes the local model of each client. The client is numbered i and the number of iterations of the current personalized federated learning method is t. S2: Client i performs the tth iteration, trains the local model by gradient descent, and uploads the trained local model to the server; S3. The server performs global aggregation on the uploaded local model to obtain a global model. S4. The server determines whether it is necessary to calculate the personalized aggregation weight based on whether the current number of iterations t is greater than the set threshold r. If not, the server sends the global model to the selected client as its initial local model and returns to step S2. If it is greater, the server proceeds to step S5. S5. The server calculates the weighted mean square error between the client's local model and the global model parameters and normalizes it to obtain the personalized aggregation weight for this iteration. S6. The server obtains the client-customized aggregated model based on the personalized aggregation weight by aggregating the client's local model and the highest p-layer network of the global model at the element level, while the remaining lower-layer networks adopt the corresponding parts of the global model; S7. The server sends the client-customized aggregation model to the selected client as its initialization local model; S8. Repeat steps S2 to S7, performing T rounds of iterations, t∈[0, T], until the local model training of each client converges; The multi-head attention-based sequence-to-sequence long short-term memory network model specifically includes: predicting the vehicle speed in a future period based on a period of historical input feature data; the model adopts an encoder-decoder architecture; in addition, the model adopts a long short-term memory network and a multi-head attention mechanism; specifically, the model first processes the input sequence through a two-layer long short-term memory network as an encoder to extract a high-dimensional representation of the time series data; then, through a customized multi-head attention layer, the sequence dependencies are captured in multiple subspaces in parallel to enhance the understanding of the complex patterns of the time series; next, the output of the multi-head attention module is sent to a two-layer long short-term memory network as a decoder to gradually construct a prediction of the future vehicle speed; finally, the output of the decoder is mapped to the predicted vehicle speed through a fully connected layer.

2. The vehicle speed prediction method for urban scenarios according to claim 1, characterized in that: In step S1, the global objective function of federated learning is defined as: in, and and is the local objective function of the i-th client; D i is the dataset of the i-th client; |D i | is the dataset size of the i-th client; Initialize the local model on the i-th client.

3. The vehicle speed prediction method for urban scenarios according to claim 2, characterized in that: In step S2, for the tth round of iterative training, the server randomly selects a subset S of N clients t Perform training and send the global model or aggregate model obtained in the previous iteration to the set S t The client in Client i performs the tth iteration and trains the local model by gradient descent as follows: Where η is the learning rate for local model training.

4. The vehicle speed prediction method for urban scenarios according to claim 3, characterized in that: In step S3, the server performs global aggregation on the uploaded local model to obtain a global model, as follows:

5. The vehicle speed prediction method for urban scenarios according to claim 4, characterized in that: In step S5, the client subset S is calculated t All local models and p higher layers of the global model: and Θ t,p The weighted mean square error of the parameters is used to obtain the model parameter difference measure M t,p , which reflects the degree of difference in model parameters between different clients, and this difference is weighted based on the number of local data samples, as follows: M t,p Each layer After normalization, the aggregation weight W is obtained t,p , making The specific method is as follows:

6. The vehicle speed prediction method for urban scenarios according to claim 5, characterized in that: In step S6, it includes: The server is based on the personalized aggregation weight W t，p , by aggregating client-side local models at the element level and the global model Θ t The highest p-layer network is used, while the remaining lower-layer networks adopt the corresponding parts of the global model to obtain the client-customized aggregation model, which is as follows: Where ⊙ is the Hadamard product, which means the elements at the same position of two matrices are multiplied; L(Θ i )yes The number of layers, and The lower layer of The elements are 0; the aggregation weight W t，p Same shape as the remaining p higher layers.

7. A vehicle speed prediction system for urban scenarios, characterized by: The system adopts the method according to any one of claims 1 to 6.

Images