Method, device, system and medium for delaying compensation of motion state information of connected vehicles

Abstract

The present application provides a method, apparatus, system and medium for delaying and compensating the motion state information of a connected vehicle. The method includes: obtaining the current motion state information and the current delay information of the sending vehicle, obtaining the historical motion state information and the corresponding historical delay information of the sending vehicle within a preset time period, inputting the current motion state information, the current delay information, the historical motion state information and the historical delay information into a long short-term memory network model, so that the long short-term memory network model outputs the predicted compensation information of the sending vehicle, and calculating the predicted compensation displacement of the sending vehicle according to the predicted compensation information. Therefore, according to the long short-term memory network model, the motion state information and the delay information of the vehicle can be modeled and predicted to compensate for the motion state information of the vehicle during the period of in-vehicle communication delay, so as to achieve the purpose of reducing the influence of in-vehicle communication delay and improving the running safety of the connected vehicle.

Description

Technical Field

[0001] The present application relates to the field of computers, and particularly to a method, device, system, and medium for compensating for the delay of the motion state information of a connected vehicle. Background Art

[0002] In the past few decades, the emergence of connected and autonomous vehicle (CAV) technology has brought new changes to the transportation system, which is conducive to significantly improving the experience of our daily driving in terms of safety, mobility, and sustainability.

[0003] Connected and autonomous vehicles equipped with communication devices can operate cooperatively. The cooperation between vehicles mainly refers to the ability to achieve all-round communication between the vehicle itself and surrounding vehicles, the environment, and the network through vehicle-to-everything (V2X) technology, including vehicle-to-vehicle (V2V), vehicle-to-infrastructure (V2I), vehicle-to-pedestrian (V2P), vehicle-to-network (V2N), etc., providing environmental perception, information interaction, and cooperative control capabilities for automotive driving and traffic management applications.

[0004] Vehicles rely on on-vehicle sensing sensors, such as cameras, radars, and lidars, to measure the states of neighboring vehicles. With the introduction of V2X communication, connected and autonomous vehicles can obtain data beyond their direct sensing range and obtain information that cannot be detected by remote sensors, which helps to improve the sensing range of connected and autonomous vehicles. However, in terms of vehicle communication technology, problems such as communication delay will inevitably be introduced, which will reduce the performance of any connected and autonomous vehicle application.

[0005] That is, after the receiving vehicle receives the motion state information sent by the sending vehicle, the sending vehicle has undergone a certain displacement during the delay time. This vehicle communication delay problem makes the information (position, speed, acceleration, etc.) of the sending vehicle obtained by the receiving vehicle not real-time, and the safety of the receiving vehicle will inevitably be affected to a certain extent. Summary of the Invention

[0006] In view of this, the purpose of the present application is to provide a method, device, system, and medium for compensating for the delay of the motion state information of a connected vehicle, which can reduce the impact of on-vehicle communication delay and improve the safety of the operation of connected vehicles.

[0007] To achieve the above purpose, the present application has the following technical solutions:

[0008] In a first aspect, an embodiment of the present application provides a method for compensating for the delay of the motion state information of a connected vehicle, including:

[0009] Obtain the current motion state information and the current delay information of the sending vehicle; the current motion state information includes the current position, the current speed, and the current acceleration; the current delay information includes the delay duration from the previous moment corresponding to the current moment to the current moment;

[0010] Obtain the historical motion state information and the corresponding historical delay information of the sending vehicle within a preset duration; the historical motion state information includes the positions at various moments, the speeds at various moments, and the accelerations at various moments; the historical delay information includes the respective delay durations from each moment to the previous moment of each moment;

[0011] Input the current motion state information, the current delay information, the historical motion state information, and the historical delay information into a long short-term memory network model, so that the long short-term memory network model outputs the predicted compensation information of the sending vehicle; the predicted compensation information includes the speed after predicted delay, the acceleration after predicted delay, and the predicted delay duration;

[0012] Calculate the predicted compensation displacement of the sending vehicle according to the predicted compensation information.

[0013] In a possible implementation manner, before inputting the current motion state information, the current delay information, the historical motion state information, and the historical delay information into the long short-term memory network model, the method further includes:

[0014] Obtain the training set of the long short-term memory network model, where the training set includes: the actual motion state information at known moments, the actual delay information at known moments, the actual motion state information at the next moment of the known moments, and the actual delay information at the next moment of the known moments;

[0015] The actual motion state information at the known moments includes the actual position, the actual speed, and the actual acceleration at the known moments; the actual delay information at the known moments includes the actual delay duration from the previous moment corresponding to the known moment to the known moment; the actual motion state information at the next moment of the known moments includes the actual position, the actual speed, and the actual acceleration at the next moment of the known moments; the actual delay information at the next moment of the known moments includes the actual delay duration from the known moment to the next moment of the known moment;

[0016] Use the training set to learn the mapping relationship between the motion state information at the known moments and the actual delay information at the known moments, and the actual motion state information at the next moment of the known moments and the actual delay information at the next moment of the known moments;

[0017] Determine the model parameters of the long short-term memory network model according to the mapping relationship.

[0018] In a possible implementation, inputting the current motion state information, the current delay information, the historical motion state information, and the historical delay information into the long short-term memory network model includes:

[0019] Encode the current motion state information, the current delay information, the historical motion state information, and the historical delay information and then input them into the long short-term memory network model;

[0020] The encoding includes performing unified standardization processing on the current motion state information, the current delay information, the historical motion state information, and the historical delay information.

[0021] In a possible implementation, the calculating the predicted compensation displacement of the sending vehicle according to the predicted compensation information includes:

[0022] The predicted compensation displacement of the sending vehicle is equal to the product of the speed after the predicted delay and the predicted delay duration.

[0023] In a second aspect, an embodiment of the present application provides a networked vehicle motion state information delay compensation device, including:

[0024] A first acquisition unit, configured to acquire the current motion state information and the current delay information of the sending vehicle; the current motion state information includes the current position, the current speed, and the current acceleration; the current delay information includes the delay duration from the previous moment corresponding to the current moment to the current moment;

[0025] A second acquisition unit, configured to acquire the historical motion state information and the corresponding historical delay information of the sending vehicle within a preset duration; the historical motion state information includes the positions at each moment, the speeds at each moment, and the accelerations at each moment; the historical delay information includes the respective delay durations from each moment to the previous moment of each moment;

[0026] An input unit, configured to input the current motion state information, the current delay information, the historical motion state information, and the historical delay information into the long short-term memory network model, so that the long short-term memory network model outputs the predicted compensation information of the sending vehicle; the predicted compensation information includes the speed after the predicted delay, the acceleration after the predicted delay, and the predicted delay duration;

[0027] A calculation unit, configured to calculate the predicted compensation displacement of the sending vehicle according to the predicted compensation information.

[0028] In a possible implementation, the device further includes:

[0029] A third acquisition unit, configured to acquire a training set of the long short-term memory network model, where the training set includes: actual motion state information at a known time, actual delay information at a known time, actual motion state information at the next time of the known time, and actual delay information at the next time of the known time;

[0030] The actual motion state information at the known time includes the actual position, actual speed, and actual acceleration at the known time; the actual delay information at the known time includes the actual delay duration from the previous time corresponding to the known time to the known time; the actual motion state information at the next time of the known time includes the actual position, actual speed, and actual acceleration at the next time of the known time; the actual delay information at the next time of the known time includes the actual delay duration from the known time to the next time of the known time;

[0031] A learning unit, configured to learn, using the training set, the mapping relationship between the motion state information at the known time and the actual delay information at the known time, and the actual motion state information at the next time of the known time and the actual delay information at the next time of the known time;

[0032] A determination unit, configured to determine the model parameters of the long short-term memory network model according to the mapping relationship.

[0033] In a possible implementation, the input unit is specifically configured to:

[0034] Encode the current motion state information, the current delay information, the historical motion state information, and the historical delay information, and then input them into the long short-term memory network model;

[0035] The encoding includes performing unified standardization processing on the current motion state information, the current delay information, the historical motion state information, and the historical delay information.

[0036] In a possible implementation, the calculation unit is specifically configured to:

[0037] The predicted compensation displacement of the sending vehicle is equal to the product of the predicted speed after the delay and the predicted delay duration.

[0038] In a third aspect, an embodiment of the present application provides a networked vehicle motion state information delay compensation system, including:

[0039] A memory, configured to store a computer program;

[0040] A processor for implementing the steps of the method for compensating the delay of the connected vehicle motion state information as described above when executing the computer program.

[0041] In a fourth aspect, an embodiment of the present application provides a computer-readable medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the method for compensating the delay of the connected vehicle motion state information as described above are implemented.

[0042] Compared with the prior art, the embodiments of the present application have the following beneficial effects:

[0043] The embodiments of the present application provide a method, device, system and medium for compensating the delay of the connected vehicle motion state information. The method includes: obtaining the current motion state information and the current delay information of the sending vehicle, where the current motion state information includes the current position, the current speed and the current acceleration, and the current delay information includes the delay duration from the previous moment corresponding to the current moment to the current moment; obtaining the historical motion state information and the corresponding historical delay information of the sending vehicle within a preset duration, where the historical motion state information includes the positions at each moment, the speeds at each moment and the accelerations at each moment, and the historical delay information includes the delay durations respectively corresponding to the previous moment from each moment to each moment; inputting the current motion state information, the current delay information, the historical motion state information and the historical delay information into a long short-term memory network model, so that the long short-term memory network model outputs the predicted compensation information of the sending vehicle, and the predicted compensation information includes the speed after predicted delay, the acceleration after predicted delay and the predicted delay duration, and calculating the predicted compensation displacement of the sending vehicle according to the predicted compensation information. Thus, according to the long short-term memory network model, the vehicle motion state information and the delay information can be modeled and predicted to compensate the vehicle motion state information during the period of in-vehicle communication delay, so as to achieve the purpose of reducing the influence of in-vehicle communication delay and improving the running safety of the connected vehicle. Description of the Drawings

[0044] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained according to these drawings without creative efforts.

[0045] Combined with the drawings and referring to the following specific embodiments, the above and other features, advantages and aspects of the various embodiments of the present disclosure will become more obvious. Throughout the drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic, and the original components and elements are not necessarily drawn to scale.

[0046] Figure 1Shows a flowchart of a method for compensating the delay of the motion state information of a connected vehicle provided by an embodiment of the present application;

[0047] Figure 2 Shows a schematic diagram of each device included in an application scenario provided by an embodiment of the present application;

[0048] Figure 3 Shows a schematic diagram of each node unit of an LSTM model provided by an embodiment of the present application;

[0049] Figure 4 Shows a flowchart of a method for compensating the delay of the motion state information of a connected vehicle provided by an embodiment of the present application;

[0050] Figure 5 Shows a schematic diagram of the relationship between historical information and predicted future information under the network architecture of an embodiment of the present application;

[0051] Figure 6 Shows a schematic diagram of the number of neurons in each layer of the LSTM model provided by an embodiment of the present application;

[0052] Figure 7 Shows a schematic diagram of a device for compensating the delay of the motion state information of a connected vehicle provided by an embodiment of the present application. Detailed implementation manners

[0053] To make the above objects, features, and advantages of the present application more obvious and understandable, the following will describe the detailed implementation manners of the present application in conjunction with the accompanying drawings.

[0054] In the following description, many specific details are set forth to facilitate a full understanding of the present application. However, the present application may also be implemented in other ways different from those described herein. Those skilled in the art can make similar generalizations without departing from the spirit of the present application. Therefore, the present application is not limited by the specific embodiments disclosed below.

[0055] As described in the background art, in the past few decades, the emergence of connected and autonomous vehicle (CAV) technology has brought new changes to the transportation system, which is conducive to a significant improvement in the experience of our daily driving in terms of safety, mobility, and sustainability.

[0056] Connected and autonomous vehicles equipped with communication devices can operate cooperatively. The cooperation between vehicles mainly refers to the ability to achieve all-round communication between the vehicle itself and surrounding vehicles, the environment, and the network through vehicle-to-everything (V2X) technology, including vehicle-to-vehicle (V2V), vehicle-to-infrastructure (V2I), vehicle-to-pedestrian (V2P), vehicle-to-network (V2N), etc., providing environmental perception, information interaction, and cooperative control capabilities for automotive driving and traffic management applications.

[0057] Vehicles rely on on-vehicle sensing sensors, such as cameras, radars, and lidars, to measure the states of neighboring vehicles. With the introduction of V2X communication, connected and autonomous vehicles can obtain data beyond their direct sensing range and acquire information that cannot be detected by remote sensors, which helps to improve the sensing range of connected and autonomous vehicles. However, in terms of vehicle communication technology, problems such as communication latency will inevitably be introduced, which will reduce the performance of any connected and autonomous vehicle application.

[0058] That is, after the receiving vehicle receives the motion state information sent by the sending vehicle, the sending vehicle has undergone a certain displacement during the delay time. This vehicle communication delay problem makes the information (position, speed, acceleration, etc.) of the sending vehicle obtained by the receiving vehicle not real-time, and the safety of the receiving vehicle will inevitably be affected to a certain extent.

[0059] The applicant's research found that regarding how to reduce the impact of in-vehicle communication latency, researchers in different directions have conducted research from different perspectives:

[0060] The communication direction focuses on reducing the impact of latency from the perspectives of communication protocol optimization or routing optimization: From the perspective of protocol optimization, a service channel time slot medium access control mechanism can be designed to improve channel utilization, increase throughput, and at the same time reduce channel latency; From the perspective of routing optimization, the routing algorithm can be improved, and information such as node position and movement speed can be used to predict the link failure time to improve performance in terms of packet end-to-end delay, transmission throughput, and message delivery rate. Although dealing with latency from a communication perspective has been done well, from the perspective of the underlying communication mechanism, the problems of latency and packet loss in V2X communication cannot be permanently eliminated.

[0061] The control direction focuses on designing controllers based on physical motion models. The controller has a feedback unit and a motion estimation unit, which can compensate for the vehicle's motion state according to real-time errors. At the same time, the communication delay is considered during the design process of the controller to explore its impact on overall stability and safety. However, this motion model is unreliable for long-term prediction, and during driving, due to the decisions made by the driver, the vehicle trajectory is often non-linear, and this model is not always perfect for dealing with non-linear situations.

[0062] For example, some examples of applying the control direction to in-vehicle communication delay compensation are using dynamic model methods for vehicle motion estimation, such as the extended Kalman filter, unscented Kalman filter, and particle filter. These filters are modified versions of the Kalman filter and are based on the first-order Markov chain, that is, the state at the current time step depends only on the state at the previous time step. Although they are useful for real-time implementation, a key limitation is that sensor data before the previous time step cannot be directly used. And the performance of filters such as the Kalman filter depends on the accuracy of the parameter matrices, especially the process noise covariance matrix Q and the measurement noise covariance matrix R. In practice, the selection of Q and R plays an important role in evaluating the Kalman filter. Since the measurement noise is device-dependent, different hardware platforms have different noise characteristics, and the theoretical derivation of these covariance matrices may not be accurate enough for all platforms.

[0063] The deep learning direction focuses on directly modeling, analyzing, and predicting the data transmitted by in-vehicle communication, which can avoid the noise modeling problem and focus on the data itself. At the same time, it can effectively solve the non-linear problems that the control direction cannot solve. Commonly used are the BP (Back Propagation) model and the RNN (Recurrent Neural Network) model. The BP neural network model processes input data by dividing it into an input layer, a hidden layer, and an output layer. The weights in the neural network are continuously updated and optimized through a feedback medium. The output data approximates the actual value with arbitrary precision to achieve the purpose of prediction. However, the problems faced by the BP neural network in the field of vehicle trajectory prediction are that the number of input features is fixed and historical trajectory data cannot be used. At the same time, the BP neural network is based on an assumption of uniform distribution and independence of input data. However, to calculate the vehicle's motion state information, the input time series data needs to be correlated. Therefore, a recursive neural network RNN can be used to design a vehicle motion state prediction model, adding memory information from the previous time step to the input at the current moment to make it have a certain memory, but the RNN has the drawback of vanishing gradients.

[0064] To solve the above technical problems, the embodiments of the present application provide a method, device, system and medium for compensating the delay of the motion state information of a connected vehicle. The method includes: obtaining the current motion state information and the current delay information of the sending vehicle, where the current motion state information includes the current position, the current speed and the current acceleration, and the current delay information includes the delay duration from the previous moment corresponding to the current moment to the current moment; obtaining the historical motion state information and the corresponding historical delay information of the sending vehicle within a preset duration, where the historical motion state information includes the positions at each moment, the speeds at each moment and the accelerations at each moment, and the historical delay information includes the delay durations respectively corresponding to the previous moment of each moment to each moment; inputting the current motion state information, the current delay information, the historical motion state information and the historical delay information into a long short-term memory network model, so that the long short-term memory network model outputs the predicted compensation information of the sending vehicle, where the predicted compensation information includes the speed after predicted delay, the acceleration after predicted delay and the predicted delay duration, and calculating the predicted compensation displacement of the sending vehicle according to the predicted compensation information. Thus, according to the long short-term memory network model, the motion state information and the delay information of the vehicle can be modeled and predicted to compensate for the motion state information of the vehicle during the period of in-vehicle communication delay, so as to achieve the purpose of reducing the influence of in-vehicle communication delay and improving the running safety of the connected vehicle.

[0065] Exemplary method

[0066] See Figure 1 As shown, it is a flowchart of a method for compensating the delay of the motion state information of a connected vehicle provided by the embodiments of the present application, including:

[0067] S101: Obtain the current motion state information and the current delay information of the sending vehicle; the current motion state information includes the current position, the current speed and the current acceleration; the current delay information includes the delay duration from the previous moment corresponding to the current moment to the current moment.

[0068] In the embodiments of the present application, the current motion state information and the current delay information of the sending vehicle can be obtained; the current motion state information can include the current position, the current speed and the current acceleration; the current delay information can include the delay duration from the previous moment corresponding to the current moment to the current moment.

[0069] Specifically, the data can come from the basic safety message set (BSM) and delay information sent by the sending vehicle through the V2X communication device. The BSM information includes key motion state information such as the current position, current speed, and current acceleration of the sending vehicle at the sending moment. This method focuses on data-driven models, making it possible to use recurrent neural networks to predict sequence-like information such as vehicle motion states. Constrained by the fact that vehicle motion state information such as road scenarios and vehicle speeds varies within a certain range, it is possible to use the vehicle's historical motion state information to predict its future motion state information.

[0070] It should be noted that the position information in the BSM is longitude, latitude, and altitude information, which does not change much within a certain period of time, and the longitude and latitude information cannot be directly applied to the model for calculation. Coordinate transformation is required, which will increase the redundancy of the algorithm. Therefore, this method does not directly use historical position information for trajectory prediction, but focuses on predicting motion state variables such as speed and acceleration for delay compensation.

[0071] For example, a possible application scenario provided by the embodiments of the present application can be, as shown in Figure 2 As shown, this scenario includes the data processing terminal 101 of the sending vehicle j; the V2X communication device 102 of the sending vehicle j; the V2X communication device 103 of the receiving vehicle i and the data processing terminal 104 of the receiving vehicle i.

[0072] The sending vehicle j sends motion state information at time t. At time t to t + 1, the sending vehicle j has displaced. At this time, the information about the sending vehicle j in the receiving vehicle i still stays at time t and before. Therefore, delay compensation is required to make up for the information about the sending vehicle j in the receiving vehicle i from time t to t + 1.

[0073] The V2X communication device 103 of the receiving vehicle i receives the BSM information about the sending vehicle j at time k sent by the V2X communication device 102 of the sending vehicle j. The BSM information contains key information such as position, speed, and acceleration, and records the delay value from the previous moment to the current moment, and then performs delay compensation for the motion information of vehicle j in the data processing terminal 104 of the receiving vehicle i.

[0074] Optionally, the current motion state information and the current delay information can be preprocessed, and the data required for subsequent neural network training can be packaged. Specifically, the packaged set of the current motion state information and the current delay information can be:

[0075] X j (t) = {v j (t), a j (t), d j(t)};

[0076] Among them, v j (t) is the current speed, a j (t) is the current acceleration, d j (t) is the time delay duration from the previous moment corresponding to the current moment to the current moment.

[0077] S102: Obtain the historical motion state information and corresponding historical delay information of the sending vehicle within a preset duration; the historical motion state information includes the positions, speeds, and accelerations at each moment; the historical delay information includes the time delay durations respectively corresponding to each moment to the previous moment of each moment.

[0078] In the embodiment of the present application, the historical motion state information and corresponding historical delay information of the sending vehicle within a preset duration can be obtained; the historical motion state information includes the positions, speeds, and accelerations at each moment; the historical delay information includes the time delay durations respectively corresponding to each moment to the previous moment of each moment.

[0079] Specifically, X j (t - 1) can be a set of the historical motion state information and corresponding historical delay information at the moment t - 1, X j (t - 1) = {v j (t - 1), a j (t - 1), d j (t - 1)}; where v j (t - 1) is the speed at the historical moment t - 1, a j (t - 1) is the acceleration at the historical moment t - 1, d j (t - 1) is the time delay duration from the historical moment t - 1 to the moment t - 2.

[0080] Thus, a total set including the current motion state information, current delay information, historical motion state information, and corresponding historical delay information can be formed:

[0081] X = {X j (t), X j (t - 1)...}.

[0082] It should be noted that the speed and acceleration information here both have horizontal and vertical directions, which are used to calculate the lateral and longitudinal displacements of the vehicle respectively. For the convenience of narration, they are both represented by one symbol here.

[0083] S103: Input the current motion state information, the current delay information, the historical motion state information, and the historical delay information into a long short-term memory network model, so that the long short-term memory network model outputs the predicted compensation information of the sending vehicle; the predicted compensation information includes the speed after predicted delay, the acceleration after predicted delay, and the predicted delay duration.

[0084] S104: Calculate the predicted compensation displacement of the sending vehicle according to the predicted compensation information.

[0085] In the embodiment of the present application, the current motion state information, the current delay information, the historical motion state information, and the historical delay information can be input into a long short-term memory network model (LSTM, Long Short-Term Memory).

[0086] Specifically, refer to Figure 3 As shown, it is a structural diagram of an LSTM unit node provided by the embodiment of the present application. At each time step t, g(x t ) is the input of this LSTM sub-unit, and h t is the output of the LSTM unit node. Combine the output h t-1 of the unit node at the previous time step with the current system input g(x t ) to form the input of the current unit. The state of each unit node is C t , which records the system memory. C t is updated at each time step. To control the information flow through the unit node, several gates are applied, including an input gate (i t ), an output gate (o t ), and a forget gate (f t ). Each gate generates an output between 0 and 1, where the value of the output is calculated by a sigmoid (σ) function. An output of 0 means that the input of the gate is completely blocked, while an output of 1 means that all the information of the input is retained in the unit node. The calculation methods of the input, output, and forget gates are as follows:

[0087] i t = σ(W i g(x t ) + U i h t-1 + b i )

[0088] o t = σ(W o g(x t ) + U o h t-1 + b o )

[0089] f t = σ(W f g(x t ) + U f h t-1 + b f );

[0090] Wherein, W i , W o and W f are the weights of three gates; U i , U o and U f are the corresponding recurrent weights; b i , b o and b f are the bias values of three gates.

[0091] Similar to the AND gate function, it combines the current input and the previous cell state C t-1 to update the cell state. The difference is that the input will be processed by a hyperbolic tangent function instead of sigmoid, which generates an output between -1 and 1:

[0092]

[0093] After the update, is multiplied by the output of the input gate and then used as the first component for updating the cell state. Another component for updating the cell state is the previous cell state, which is processed by the forget gate to determine how to use past data. For these two components, the cell state at time t will be updated to:

[0094]

[0095] The output of the cell h t to be used at t + 1 is calculated by multiplying the output gate with the tanh function of the current cell state:

[0096] h t = o t tanh(c t ).

[0097] In a possible implementation, to improve the accuracy of model prediction, a semi-supervised learning strategy can be adopted to obtain the training set of the long short-term memory network model. The training set includes: the actual motion state information at known times, the actual delay information at known times, the actual motion state information at the next moment of known times, and the actual delay information at the next moment of known times.

[0098] The actual motion state information at a known time includes the actual position, actual velocity, and actual acceleration at the known time; the actual delay information at a known time includes the actual delay duration from the previous time corresponding to the known time to the known time; the actual motion state information at the next time of the known time includes the actual position, actual velocity, and actual acceleration at the next time of the known time; the actual delay information at the next time of the known time includes the actual delay duration from the known time to the next time of the known time;

[0099] Use the training set to learn the mapping relationship between the motion state information and actual delay information at a known time, and the actual motion state information and actual delay information at the next time of the known time;

[0100] Determine the model parameters of the long short-term memory network model according to the mapping relationship, so as to use the state information and delay information at several historical times as training data, and the state information at the next time as data labels for training to improve the prediction accuracy of the model.

[0101] In a possible implementation, see Figure 4 As shown, in order to improve the prediction stability and performance of the long short-term memory network, the current motion state information, current delay information, historical motion state information, and historical delay information can be encoded and then input into the long short-term memory network model.

[0102] Specifically, the encoding includes uniformly standardizing the current motion state information, current delay information, historical motion state information, and historical delay information. Subsequently, training and using the network model helps to improve the stability and performance of the network, and the encoding can be performed using an encoder.

[0103] In addition, a decoder can be used to perform inverse standardization of the data according to the relevant encoding of the encoder to achieve the purpose of data scaling.

[0104] Optionally, the following standard is adopted:

[0105] where is the nth component of the input data, such as the delay, velocity, or acceleration at time t, that is, x t，n is the standardized input, μ N , σ N are the mean and variance obtained from the total samples.

[0106] See Figure 5 As shown, it is a relationship diagram of historical information and predicted future information under the method network architecture provided by the embodiments of the present application. For each time T t-h+1 to T tThe Input at each moment is encoded and then input into the LSTM model, and finally decoded to output, so as to predict the predicted compensation information Output at the future T t+1 moments.

[0107] The length of the input sequence, that is, the historical time step, is an important factor affecting the prediction performance. According to the preliminary experiments of the inventors, a length of 5 to 15 is appropriate.

[0108] It should be noted that the sequences used for neural network training in this method are of equal step length, but the motion state information in the sequences is the result of changes after several unequal delay values, which is also one of the reasons for not directly using position information for trajectory prediction.

[0109] Specifically, for the LSTM model, see Figure 6 As shown, from the input layer (Input Layer) to the fully connected layer (Full connect Layer), then to the LSTM layer, then to the LSTM layer, and then to the fully connected layer, and finally perform regression prediction (Regression). The numbers in parentheses are the number of neurons in each sub-layer.

[0110] Optionally, taking uniformly variable motion as an example, a detailed description of the compensated motion state information is made.

[0111] The predicted state information after delay is: X j (t + 1) = {v j (t + 1), a j (t + 1), d j (t + 1)}

[0112] Among them, X j (t + 1) is the information set after delay, v j (t + 1) is the predicted speed after delay, a j (t + 1) is the predicted acceleration, d j (t + 1) is the predicted delay value from time t to time t + 1.

[0113] That is, the predicted compensation displacement of the sending vehicle can be calculated based on the predicted compensation information. Specifically, the predicted compensation displacement Δp of the sending vehicle is equal to the product of the speed after predicted delay and the predicted delay duration.

[0114] Δp = v j (t + 1) * d j (t + 1);

[0115] The speed after predicted delay: v j (t + 1) = a j (t + 1) * d j (t + 1).

[0116] An embodiment of the present application provides a method for compensating the delay of the motion state information of a connected vehicle. The method includes: obtaining the current motion state information and the current delay information of the sending vehicle, where the current motion state information includes the current position, the current speed, and the current acceleration, and the current delay information includes the delay duration from the previous moment corresponding to the current moment to the current moment; obtaining the historical motion state information and the corresponding historical delay information of the sending vehicle within a preset duration, where the historical motion state information includes the positions at each moment, the speeds at each moment, and the accelerations at each moment, and the historical delay information includes the respective delay durations from each moment to the previous moment of each moment; inputting the current motion state information, the current delay information, the historical motion state information, and the historical delay information into a long short-term memory network model, so that the long short-term memory network model outputs the predicted compensation information of the sending vehicle, where the predicted compensation information includes the speed after predicted delay, the acceleration after predicted delay, and the predicted delay duration, and calculating the predicted compensation displacement of the sending vehicle according to the predicted compensation information. Thus, according to the long short-term memory network model, the motion state information and the delay information of the vehicle can be modeled and predicted to compensate for the motion state information of the vehicle during the period of in-vehicle communication delay, so as to achieve the purpose of reducing the impact of in-vehicle communication delay and improving the running safety of the connected vehicle.

[0117] Exemplary device

[0118] See Figure 2 As shown in the figure, an apparatus for compensating the delay of the motion state information of a connected vehicle provided by an embodiment of the present application includes:

[0119] A first acquisition unit 201, configured to acquire the current motion state information and the current delay information of the sending vehicle; the current motion state information includes the current position, the current speed, and the current acceleration; the current delay information includes the delay duration from the previous moment corresponding to the current moment to the current moment;

[0120] A second acquisition unit 202, configured to acquire the historical motion state information and the corresponding historical delay information of the sending vehicle within a preset duration; the historical motion state information includes the positions at each moment, the speeds at each moment, and the accelerations at each moment; the historical delay information includes the respective delay durations from each moment to the previous moment of each moment;

[0121] An input unit 203, configured to input the current motion state information, the current delay information, the historical motion state information, and the historical delay information into a long short-term memory network model, so that the long short-term memory network model outputs the predicted compensation information of the sending vehicle; the predicted compensation information includes the speed after predicted delay, the acceleration after predicted delay, and the predicted delay duration;

[0122] A calculation unit 204, configured to calculate a predicted compensation displacement of the sending vehicle according to the predicted compensation information.

[0123] In a possible implementation, the apparatus further includes:

[0124] A third acquisition unit, configured to acquire a training set of the long short-term memory network model, where the training set includes: actual motion state information at a known time, actual delay information at a known time, actual motion state information at the next moment of the known time, and actual delay information at the next moment of the known time;

[0125] The actual motion state information at the known time includes the actual position, actual speed, and actual acceleration at the known time; the actual delay information at the known time includes the actual delay duration from the previous moment corresponding to the known time to the known time; the actual motion state information at the next moment of the known time includes the actual position, actual speed, and actual acceleration at the next moment of the known time; the actual delay information at the next moment of the known time includes the actual delay duration from the known time to the next moment of the known time;

[0126] A learning unit, configured to learn a mapping relationship between the motion state information at the known time and the actual delay information at the known time, and the actual motion state information at the next moment of the known time and the actual delay information at the next moment of the known time by using the training set;

[0127] A determination unit, configured to determine model parameters of the long short-term memory network model according to the mapping relationship.

[0128] In a possible implementation, the input unit is specifically configured to:

[0129] Encode the current motion state information, the current delay information, the historical motion state information, and the historical delay information, and then input the encoded information into the long short-term memory network model;

[0130] The encoding includes uniformly standardizing the current motion state information, the current delay information, the historical motion state information, and the historical delay information.

[0131] In a possible implementation, the calculation unit is specifically configured to:

[0132] The predicted compensation displacement of the sending vehicle is equal to the product of the speed after the predicted delay and the predicted delay duration.

[0133] An embodiment of the present application provides a device for compensating the delay of the motion state information of a connected vehicle. The method using this device includes: obtaining the current motion state information and the current delay information of the sending vehicle. The current motion state information includes the current position, the current speed, and the current acceleration. The current delay information includes the delay duration from the previous moment corresponding to the current moment to the current moment; obtaining the historical motion state information and the corresponding historical delay information of the sending vehicle within a preset duration. The historical motion state information includes the positions at each moment, the speeds at each moment, and the accelerations at each moment. The historical delay information includes the delay durations respectively corresponding to the previous moment of each moment to each moment; inputting the current motion state information, the current delay information, the historical motion state information, and the historical delay information into a long short-term memory network model, so that the long short-term memory network model outputs the predicted compensation information of the sending vehicle. The predicted compensation information includes the speed after predicted delay, the acceleration after predicted delay, and the predicted delay duration. The predicted compensation displacement of the sending vehicle is calculated according to the predicted compensation information. Thus, according to the long short-term memory network model, the motion state information and the delay information of the vehicle can be modeled and predicted to compensate for the motion state information of the vehicle during the period of in-vehicle communication delay, so as to achieve the purpose of reducing the impact of in-vehicle communication delay and improving the safety of the operation of connected vehicles.

[0134] On the basis of the above embodiment, an embodiment of the present application provides a system for compensating the delay of the motion state information of a connected vehicle, including:

[0135] A memory for storing a computer program;

[0136] A processor for implementing the steps of the above method for compensating the delay of the motion state information of a connected vehicle when executing the computer program.

[0137] On the basis of the above embodiment, an embodiment of the present application further provides a computer-readable medium, on which a computer program is stored, and the computer program realizes the steps of the above method for compensating the delay of the motion state information of a connected vehicle when being processed and executed.

[0138] It should be noted that the computer-readable medium described above in the present disclosure can be a computer-readable signal medium, a computer-readable storage medium, or any combination of the two. A computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples of the computer-readable storage medium can include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present disclosure, a computer-readable storage medium can be any tangible medium that contains or stores a program, and this program can be used by or in combination with an instruction execution system, apparatus, or device. In the present disclosure, a computer-readable signal medium can include a data signal propagated in a baseband or as part of a carrier wave, which carries computer-readable program code. Such a propagated data signal can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. A computer-readable signal medium can also be any computer-readable medium other than a computer-readable storage medium, and this computer-readable signal medium can send, propagate, or transmit a program for use by or in combination with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted by any appropriate medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0139] The above computer-readable medium can be included in the above system; it can also exist separately without being assembled into the system.

[0140] In particular, according to the embodiments of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments of the present disclosure include a computer program product, which includes a computer program carried on a non-transitory computer-readable medium, and this computer program contains program code for executing the method shown in the flowchart.

[0141] Each embodiment in this specification is described in a progressive manner. For the same or similar parts among the embodiments, reference can be made to each other. Each embodiment focuses on the differences from other embodiments. In particular, for the device embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and for the relevant parts, reference can be made to the partial description of the method embodiments.

[0142] The above are only the preferred embodiments of the present application. Although the present application has been disclosed above with preferred embodiments, it is not intended to limit the present application. Any person skilled in the art can make many possible changes and modifications to the technical solution of the present application, or modify it into an equivalent embodiment with equivalent changes, without departing from the scope of the technical solution of the present application. Therefore, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present application without departing from the content of the technical solution of the present application still fall within the scope of protection of the technical solution of the present application.

Claims

1. A method for compensating the delay of the motion state information of a connected vehicle, characterized in that, it includes: Obtain the current motion state information and current delay information of the sending vehicle; The current motion state information includes the current position, current speed and current acceleration; The current delay information includes the time difference between the moment when the receiving vehicle receives the current motion state information and the moment when the sending vehicle sends the current motion state information; Obtain the historical motion state information and corresponding historical delay information of the sending vehicle within a preset duration; the historical motion state information includes the positions, speeds and accelerations at each moment; the historical delay information includes the time difference between the moment when the receiving vehicle receives the current motion state information each time and the moment when the sending vehicle sends the current motion state information this time; Input the current motion state information, the current delay information, the historical motion state information and the historical delay information into a long short-term memory network model, so that the long short-term memory network model outputs the predicted compensation information of the sending vehicle; the predicted compensation information includes the predicted speed after delay and the predicted delay duration; Obtain the predicted compensation displacement of the sending vehicle according to the product of the predicted speed after delay and the predicted delay duration.

2. The method according to claim 1, characterized in that, Before inputting the current motion state information, the current delay information, the historical motion state information and the historical delay information into the long short-term memory network model, the method further includes: Obtain the training set of the long short-term memory network model, and the training set includes: the actual motion state information at a known moment, the actual delay information at a known moment, the actual motion state information at the next moment of the known moment and the actual delay information at the next moment of the known moment; The actual motion state information at the known moment includes the actual position, actual speed and actual acceleration at the known moment; the actual delay information at the known moment includes the actual delay duration from the previous moment corresponding to the known moment to the known moment; the actual motion state information at the next moment of the known moment includes the actual position, actual speed and actual acceleration at the next moment of the known moment; the actual delay information at the next moment of the known moment includes the actual delay duration from the known moment to the next moment of the known moment; Use the training set to learn the mapping relationship between the motion state information at the known moment and the actual delay information at the known moment, and the actual motion state information at the next moment of the known moment and the actual delay information at the next moment of the known moment; Determine the model parameters of the long short-term memory network model according to the mapping relationship.

3. The method according to claim 1, characterized in that, Inputting the current motion state information, the current delay information, the historical motion state information and the historical delay information into the long short-term memory network model includes: Encode the current motion state information, the current delay information, the historical motion state information and the historical delay information and then input them into the long short-term memory network model; The encoding includes uniformly standardizing the current motion state information, the current delay information, the historical motion state information, and the historical delay information.

4. A delay compensation device for the motion state information of a connected vehicle, characterized in that, it includes: A first acquisition unit for acquiring the current motion state information and the current delay information of the sending vehicle; The current motion state information includes the current position, the current speed, and the current acceleration; the current delay information includes the time difference between the moment when the receiving vehicle receives the current motion state information and the moment when the sending vehicle sends the current motion state information; A second acquisition unit for acquiring the historical motion state information and the corresponding historical delay information of the sending vehicle within a preset time period; the historical motion state information includes the positions, speeds, and accelerations at each moment; the historical delay information includes the time difference between the moment when the receiving vehicle receives the current motion state information each time and the moment when the sending vehicle sends the current motion state information this time; An input unit for inputting the current motion state information, the current delay information, the historical motion state information, and the historical delay information into a long short-term memory network model so that the long short-term memory network model outputs the predicted compensation information of the sending vehicle; the predicted compensation information includes the speed after predicted delay, the acceleration after predicted delay, and the predicted delay duration; A calculation unit for calculating the predicted compensation displacement of the sending vehicle according to the predicted compensation information. Specifically, the calculation unit is configured to: obtain the predicted compensation displacement of the sending vehicle according to the product of the speed after predicted delay and the predicted delay duration.

5. The device according to claim 4, characterized in that, the device further includes: A third acquisition unit for acquiring the training set of the long short-term memory network model. The training set includes: the actual motion state information at known moments, the actual delay information at known moments, the actual motion state information at the next moment of the known moments, and the actual delay information at the next moment of the known moments; The actual motion state information at the known moments includes the actual position, actual speed, and actual acceleration at the known moments; the actual delay information at the known moments includes the actual delay duration from the previous moment corresponding to the known moment to the known moment; the actual motion state information at the next moment of the known moments includes the actual position, actual speed, and actual acceleration at the next moment of the known moments; the actual delay information at the next moment of the known moments includes the actual delay duration from the known moment to the next moment of the known moment; A learning unit for learning the mapping relationship between the motion state information and the actual delay information at the known moments and the actual motion state information and the actual delay information at the next moment of the known moments by using the training set; A determination unit for determining the model parameters of the long short-term memory network model according to the mapping relationship.

6. The device according to claim 4, characterized in that, the input unit is specifically configured to: Encode the current motion state information, the current delay information, the historical motion state information, and the historical delay information, and then input them into the long short-term memory network model; The encoding includes uniformly standardizing the current motion state information, the current delay information, the historical motion state information, and the historical delay information.

7. A connected vehicle motion state information delay compensation system Characterized in that It includes: A memory for storing computer programs; A processor for implementing the steps of the connected vehicle motion state information delay compensation method according to any one of claims 1-4 when executing the computer program.

8. A computer-readable medium Characterized in that The computer-readable medium stores a computer program, and when the computer program is processed and executed, the steps of the connected vehicle motion state information delay compensation method according to any one of claims 1-4 are implemented.

Images