Power output data prediction method and device for vehicle in multiple driving scenes

Abstract

The invention relates to the technical field of vehicle load prediction, and discloses a power output data prediction method and device for a vehicle in multiple driving scenes. The method is based on a power output power predicted value predicted by a first sub-network, a torque demand predicted value predicted by a second sub-network, a long-term power demand predicted value predicted by a third sub-network and an energy recovery time predicted value, and the first sub-network, the second sub-network and the third sub-network form a multi-layer network structure. Obtaining a sub-network corresponding to the optimal state of the vehicle in the current driving scene; and obtaining a prediction value output by the sub-network corresponding to the optimal state, and taking the prediction value as an optimal prediction result in the current driving scene. According to the method, the most suitable sub-network can be dynamically selected, compared with a traditional static model, the method has higher flexibility and adaptability, the prediction result output by the optimal sub-network can be obtained through matching, and the prediction precision of the vehicle power output data is further improved.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle power output data prediction, and in particular to a method and device for predicting power output data of a vehicle in multiple driving scenarios. Background Art

[0002] In the intelligent transportation and automotive industries, drive system optimization technology plays a relatively large role in improving the energy efficiency, power response and overall driving performance of vehicles. And with the development of intelligent driving technology, the performance of the vehicle drive system not only affects the vehicle's power output and fuel / battery efficiency, but is also directly related to driving safety and driving experience. Therefore, how to effectively optimize the vehicle's drive system has become an important direction of current research.

[0003] At present, the traditional vehicle drive system optimization method usually sets preset rules or fixed parameters, but this method is difficult to adapt to dynamic and complex driving scenarios. Vehicles need to deal with various driving scenarios during driving, such as urban congestion, high-speed driving, ramp driving, etc., which makes it impossible for a single drive system optimization strategy to take into account different needs. However, in order to improve the adaptability of the drive system, modern vehicles are usually equipped with a variety of sensors for real-time collection of vehicle operating status, road conditions and environmental data. These data also provide a data-driven basis for drive system optimization to a certain extent. However, when predicting power output data in different driving scenarios, related technologies do not distinguish between complex and diverse test data, and use the same neural network structure to predict various test data at the same time, which can easily lead to deviations in the prediction results of power output data. Summary of the invention

[0004] In view of this, the present invention provides a method and device for predicting power output data of a vehicle in multiple driving scenarios, so as to solve the problem that using the same neural network structure to simultaneously predict various detection data easily leads to deviations in the prediction results of power output data.

[0005] According to a first aspect, an embodiment of the present disclosure provides a method for predicting power output data of a vehicle in multiple driving scenarios, the method comprising:

[0006] Obtain the current detection data, historical average power demand data, and historical energy recovery time data of the vehicle in the current driving scenario. The current detection data includes: vehicle operation data, vehicle road condition data, traffic data, and driving environment data;

[0007] The first sub-network predicts the power output power prediction value corresponding to the vehicle road condition data, the second sub-network predicts the traffic data and the torque demand prediction value corresponding to the driving environment data in combination with the power output power prediction value, and the third sub-network predicts the long-term power demand prediction value corresponding to the historical average power demand data and the energy recovery time prediction value corresponding to the historical energy recovery time data, respectively. The first sub-network, the second sub-network and the third sub-network form a multi-layer network structure;

[0008] Based on the power output power prediction value, torque demand prediction value, long-term power demand prediction value, and energy recovery time prediction value, obtain the sub-network corresponding to the optimal state of the vehicle in the current driving scenario;

[0009] Get the predicted value of the sub-network output corresponding to the optimal state, and use the predicted value as the optimal prediction result in the current driving scenario.

[0010] In some optional implementations, predicting the first load prediction value corresponding to the vehicle road condition data by the first subnetwork includes:

[0011] Obtain speed, acceleration, and engine speed from vehicle traffic data;

[0012] Based on the speed, acceleration and engine speed, the power output power prediction value corresponding to the vehicle road condition data predicted by the first sub-network is calculated, wherein the power output power prediction value is calculated by the following formula;

[0013] P pred (t) = f conv (v(t),a(t),n(t))

[0014] Among them, P pred (t) is the predicted value of power output, f conv (v(t) is the convolution operation factor, v(t) is the velocity, a(t) is the acceleration, and n(t) is the engine speed.

[0015] In some optional implementations, the torque demand prediction value corresponding to the vehicle road condition data, traffic data, and driving environment data is predicted by the second sub-network in combination with the power output power prediction value, including:

[0016] Obtain slope from vehicle road condition data, obtain traffic flow from traffic data, and obtain temperature from driving environment data;

[0017] Based on the power output power prediction value, slope, traffic flow, and temperature, the torque demand prediction value corresponding to the traffic data and driving environment data predicted by the second sub-network is calculated, wherein the torque demand prediction value is calculated by the following formula;

[0018] B pred(t) = f merge (P pred (t),θ(t),T(t),T f (t)

[0019] Among them, B pred (t) is the predicted value of torque demand, f merge is the feature fusion factor, P pred (t) is the predicted value of power output, θ(t) is the slope, T(t) is the temperature, T f (t) is the traffic flow.

[0020] In some optional implementations, respectively predicting a long-term power demand prediction value corresponding to the historical average power demand data and an energy recovery time prediction value corresponding to the historical energy recovery time data by a third sub-network includes:

[0021] Based on the historical average power demand data and the historical energy recovery time data, the long-term power demand forecast value and the energy recovery time forecast value are calculated, wherein the long-term power demand forecast value and the energy recovery time forecast value are calculated by the following formula;

[0022] P long (t),E long (t) = f LSTM (P avg (t),E rec (t)

[0023] Among them, P long (t) is the long-term power demand forecast value, E pred (t) is the predicted value of energy recovery time, f LSTM is the calculation factor of the third sub-network, P avg (t) is the historical average power demand data, E rec (t) is the historical energy recovery time data.

[0024] In some optional implementations, based on the power output power prediction value, the torque demand prediction value, the long-term power demand prediction value, and the energy recovery time prediction value, the subnetwork corresponding to the optimal state of the vehicle in the current driving scenario is obtained, and the following formula is used for execution:

[0025]

[0026] Among them, S(t) is the state feature vector corresponding to the optimal state of the vehicle by real-time monitoring of the current detection data of the vehicle in the current driving scene, and f i(X(t), Y(t)) is the output accuracy of the i-th sub-network corresponding to the optimal state in the current driving scenario, X(t) is the input feature of the i-th sub-network, and Y(t) is the predicted value output by the i-th sub-network.

[0027] In some optional implementations, the current driving scenario includes a city congestion scenario, a high-speed driving scenario, or a ramp driving scenario, and further includes:

[0028] The current detection data in the urban congestion scenario, high-speed driving scenario, or slope driving scenario is weighted.

[0029] According to a second aspect, an embodiment of the present disclosure provides a device for predicting power output data of a vehicle in multiple driving scenarios, the device comprising:

[0030] The first acquisition module is used to acquire the current detection data, historical average power demand data, and historical energy recovery time data of the vehicle in the current driving scenario. The current detection data includes: vehicle operation data, vehicle road condition data, traffic data, and driving environment data. The current driving scenario includes a city congestion scenario, a high-speed driving scenario, or a slope driving scenario;

[0031] A data prediction module is used to predict the power output power prediction value corresponding to the vehicle road condition data through the first sub-network, predict the torque demand prediction value corresponding to the traffic data and the driving environment data through the second sub-network combined with the power output power prediction value, and respectively predict the long-term power demand prediction value corresponding to the historical average power demand data and the energy recovery time prediction value corresponding to the historical energy recovery time data through the third sub-network, wherein the first sub-network, the second sub-network and the third sub-network constitute a multi-layer network structure;

[0032] The second acquisition module is used to acquire the sub-network corresponding to the optimal state of the vehicle in the current driving scenario based on the power output power prediction value, the torque demand prediction value, the long-term power demand prediction value, and the energy recovery time prediction value;

[0033] The third acquisition module is used to obtain the predicted value of the sub-network output corresponding to the optimal state, and use the predicted value as the optimal prediction result in the current driving scenario.

[0034] In a third aspect, an embodiment of the present disclosure provides a computer device, including:

[0035] The memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the method for predicting power output data of a vehicle in multiple driving scenarios according to the first aspect or any corresponding embodiment thereof by executing the computer instructions.

[0036] In a fourth aspect, an embodiment of the present disclosure provides a computer-readable storage medium having computer instructions stored thereon, the computer instructions being used to enable a computer to execute the method for predicting power output data of a vehicle in multiple driving scenarios of the first aspect or any corresponding embodiment thereof.

[0037] In a fifth aspect, an embodiment of the present disclosure provides a computer program product, including computer instructions, which are used to enable a computer to execute the method for predicting power output data of a vehicle in multiple driving scenarios according to the above-mentioned first aspect or any corresponding embodiment.

[0038] The technical solution of the present invention has the following advantages:

[0039] The present invention relates to the technical field of vehicle power output data prediction, and discloses a method and device for predicting power output data of a vehicle in multiple driving scenarios. The present invention is based on the power output power prediction value predicted by the first sub-network, the torque demand prediction value predicted by the second sub-network, the long-term power demand prediction value predicted by the third sub-network, and the energy recovery time prediction value, wherein the first sub-network, the second sub-network, and the third sub-network constitute a multi-layer network structure, and obtain the sub-network corresponding to the optimal state of the vehicle in the current driving scenario; obtain the predicted value of the sub-network output corresponding to the optimal state, and use the predicted value as the optimal prediction result in the current driving scenario. The present invention can dynamically select the most suitable sub-network, and compared with the traditional static model, it has higher flexibility and adaptability, can match the prediction result of the optimal sub-network output, and further improve the prediction accuracy of the vehicle power output data. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0041] Figure 1 is a flow chart of a method for predicting power output data of a vehicle in multiple driving scenarios according to an embodiment of the present invention;

[0042] Figure 2 is a simplified block diagram of a multi-layer network structure according to an embodiment of the present invention;

[0043] Figure 3 is a flow chart of another method for predicting power output data of a vehicle in multiple driving scenarios according to an embodiment of the present invention;

[0044] Figure 4is a flow chart of another method for predicting power output data of a vehicle in multiple driving scenarios according to an embodiment of the present invention;

[0045] Figure 5 is a simplified block diagram of a vehicle collecting current vehicle detection data in real time under multiple driving scenarios according to an embodiment of the present invention;

[0046] Figure 6 is a flow chart of a method for predicting power output data of a vehicle in multiple driving scenarios according to another embodiment of the present invention;

[0047] Figure 7 is a structural block diagram of a power output data prediction device for a vehicle in multiple driving scenarios according to an embodiment of the present invention;

[0048] Figure 8 It is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0049] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0050] According to an embodiment of the present invention, an embodiment of a method for predicting power output data of a vehicle in multiple driving scenarios is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in an order different from that shown here.

[0051] In this embodiment, a method for predicting power output data of a vehicle in multiple driving scenarios is provided, which can be used in computer devices such as mobile phones, tablet computers, desktop computers, portable notebooks, servers, etc. Figure 1 is a flow chart of a method for predicting power output data of a vehicle in multiple driving scenarios according to an embodiment of the present invention, the process comprising the following steps:

[0052] Step S101, obtaining current detection data, historical average power demand data, and historical energy recovery time data of the vehicle in the current driving scenario, where the current detection data includes: vehicle operation data, vehicle road condition data, traffic data, and driving environment data.

[0053] Specifically, the vehicle operation data includes: acceleration, speed, engine speed, among which, acceleration, among which, v(t) is the vehicle speed, t is the time, a(t) is the acceleration, which is detected by the acceleration sensor, and the speed is detected by the vehicle speed sensor. Vehicle road condition data includes: slope, which can be detected by the camera, lidar sensor, and millimeter wave radar sensor. Traffic data includes: traffic flow, Among them, N vehicles (t) is the unit road area A at the current moment lane The number of vehicles on the vehicle, the traffic flow data is detected by a traffic flow sensor. The driving environment data includes: temperature, the temperature is detected by a temperature sensor.

[0054] Furthermore, the method for predicting power output data of a vehicle in multiple driving scenarios in the embodiment of the present disclosure supports driving in multiple driving environments, thereby meeting different driving needs of driving users, and can flexibly predict power output data in multiple driving scenarios.

[0055] Step S102, predicting the power output power prediction value corresponding to the vehicle road condition data through the first sub-network, predicting the traffic data and the torque demand prediction value corresponding to the driving environment data through the second sub-network combined with the power output power prediction value, and predicting the long-term power demand prediction value corresponding to the historical average power demand data and the energy recovery time prediction value corresponding to the historical energy recovery time data through the third sub-network, respectively. The first sub-network, the second sub-network and the third sub-network constitute a multi-layer network structure.

[0056] Specifically, Figure 2 As shown in Figure 1, it is a simple block diagram of a multi-layer network structure. Figure 2 The first subnetwork, the second subnetwork and the third subnetwork constitute a multi-layer network structure, wherein the first subnetwork is used as the bottom layer network to process basic data, the second subnetwork is used as the middle layer network to integrate road condition data, environmental data and traffic data, and the third subnetwork is used as the high layer network to process historical data.

[0057] In some optional implementations, predicting the first load prediction value corresponding to the vehicle road condition data by the first subnetwork includes:

[0058] Step a1, obtaining speed, acceleration, and engine speed from vehicle road condition data.

[0059] Step a2, based on the speed, acceleration and engine speed, calculate the power output power prediction value corresponding to the vehicle road condition data predicted by the first sub-network, wherein the power output power prediction value is calculated by the following formula;

[0060] P pred(t) = f conv (v(t),a(t),n(t))

[0061] Among them, P pred (t) is the predicted value of power output, f conv (v(t) is the convolution operation factor, v(t) is the velocity, a(t) is the acceleration, and n(t) is the engine speed.

[0062] Specifically, the speed, acceleration, and engine speed are extracted through multiple convolutional layers of the first sub-network, and then a preliminary power output power prediction value is generated through the fully connected layer. The power output prediction value serves as the first prediction task.

[0063] In some optional implementations, the torque demand prediction value corresponding to the vehicle road condition data, traffic data, and driving environment data is predicted by the second sub-network in combination with the power output power prediction value, including:

[0064] Step b1, obtaining the slope from the vehicle road condition data, obtaining the traffic flow from the traffic data, and obtaining the temperature from the driving environment data.

[0065] Step b2, based on the power output power prediction value, slope, traffic flow, and temperature, calculate the torque demand prediction value corresponding to the traffic data predicted by the second sub-network and the driving environment data, wherein the torque demand prediction value is calculated by the following formula;

[0066] B pred (t) = f merge (P pred (t),θ(t),T(t),T f (t)

[0067] Among them, B pred (t) is the predicted value of torque demand, f merge is the feature fusion factor, P pred (t) is the predicted value of power output, θ(t) is the slope, T(t) is the temperature, T f (t) is the traffic flow.

[0068] Specifically, the second sub-network combines the power output power prediction value, slope, traffic flow, and temperature output by the first sub-network to generate a torque demand prediction value corresponding to the second sub-network predicted traffic data and driving environment data.

[0069] In some optional implementations, respectively predicting a long-term power demand prediction value corresponding to the historical average power demand data and an energy recovery time prediction value corresponding to the historical energy recovery time data by a third sub-network includes:

[0070] Step c1, based on the historical average power demand data and the historical energy recovery time data, calculate the long-term power demand forecast value and the energy recovery time forecast value, wherein the long-term power demand forecast value and the energy recovery time forecast value are calculated by the following formula.

[0071] P long (t),E long (t) = f LSTM (P avg (t),E rec (t)

[0072] Among them, P long (t) is the long-term power demand forecast value, E pred (t) is the predicted value of energy recovery time, f LSTM is the calculation factor of the third sub-network, P avg (t) is the historical average power demand data, E rec (t) is the historical energy recovery time data.

[0073] Specifically, the third sub-network performs long-term power demand prediction based on historical data. The embodiment of the present disclosure mainly performs time series modeling based on the long short-term memory network (LSTM), and the input is the historical average power demand data P avg (t) and historical energy recovery time data E rec (t), the output is the long-term power demand forecast value P long (t) and the predicted energy recovery time E pred (t).

[0074] The disclosed embodiment predicts a variety of data in the current driving scenario through a plurality of different sub-networks, which is beneficial to improving the accuracy of the prediction results of the vehicle power output data.

[0075] Step S103, based on the power output power prediction value, the torque demand prediction value, the long-term power demand prediction value, and the energy recovery time prediction value, obtain the sub-network corresponding to the optimal state of the vehicle in the current driving scenario.

[0076] Step S104, obtaining the predicted value output by the sub-network corresponding to the optimal state, and using the predicted value as the optimal prediction result in the current driving scenario.

[0077] In some optional implementations, based on the power output power prediction value, the torque demand prediction value, the long-term power demand prediction value, and the energy recovery time prediction value, the subnetwork corresponding to the optimal state of the vehicle in the current driving scenario is obtained, and the following formula is used for execution:

[0078]

[0079] Among them, S(t) is the state feature vector corresponding to the optimal state of the vehicle by real-time monitoring of the current detection data of the vehicle in the current driving scene, and f i (X(t), Y(t)) is the output accuracy of the i-th sub-network corresponding to the optimal state in the current driving scenario, X(t) is the input feature of the i-th sub-network, and Y(t) is the predicted value output by the i-th sub-network.

[0080] The disclosed embodiment dynamically selects the most suitable sub-network by monitoring the real-time detection data of the vehicle in the current driving scenario (such as vehicle speed, acceleration, environmental data, and traffic data). Compared with the traditional static model, it has higher flexibility and adaptability, can match the prediction results of the optimal sub-network output, and further improves the prediction accuracy of the vehicle power output data.

[0081] In some optional implementations, the current driving scene includes a city congestion scene, a high-speed driving scene, or a ramp driving scene, and further includes: weighted processing of current detection data in the city congestion scene, the high-speed driving scene, or the ramp driving scene.

[0082] Specifically, for the urban congestion scenario, the main focus is on dealing with the frequent starting and stopping of vehicles, focusing on predicting the braking frequency and energy consumption at low speeds, and performing weighted processing on the current detection data in the urban congestion scenario, which is executed through the following formula:

[0083] E city (t)=α1·v(t)+α2·a(t)+α3·f b (t)+α4·p th (t)

[0084] Among them, α1 is the weight corresponding to the speed, α2 is the weight corresponding to the acceleration, α3 is the weight corresponding to the braking frequency, α4 is the weight corresponding to the braking operation threshold, v(t) is the speed, a(t) is the acceleration, f b (t) is the braking frequency, p th (t) is the braking operation threshold, E city (t) is the energy consumption demand value under urban congestion scenarios.

[0085] By weighting the current detection data in the urban congestion scenario, it is convenient to timely process the power output of the vehicle in the urban congestion scenario. For example, in the urban congestion scenario, the focus is on the low-speed braking of the vehicle. b (t) exceeds p th(t) and v(t) belongs to the low speed range, and it is determined to belong to the urban congestion scene. By executing the above steps S101 to S104, the predicted value of the sub-network output corresponding to the optimal state is obtained, and the predicted value is used as the optimal prediction result in the urban congestion scene, so as to handle the frequent braking and starting in the urban congestion scene.

[0086] Specifically, for high-speed driving scenarios, the main processing is the load prediction of high-speed stable driving, focusing on vehicle speed, engine speed, power output, and energy recovery time. The current detection data in the high-speed driving scenario is weighted and executed through the following formula:

[0087] P avg1 (t)=β1·v(t)+β2·n(t)+β3·P(t)+β4·E h (t)

[0088] Among them, β1 is the weight corresponding to the speed, β2 is the weight corresponding to the engine speed, β3 is the weight corresponding to the power output, β4 is the weight corresponding to the energy recovery time, v(t) is the speed, n(t) is the engine speed, P(t) is the power output, E h (t) is the energy recovery time, P avg1 (t) is the power output demand value in this high-speed driving scenario.

[0089] By weighting the current detection data in the high-speed driving scenario, it is convenient to timely process the power output of the vehicle in the high-speed driving scenario. For example, when v(t) is continuously higher than the set value and n(t) meets the preset range, it is determined to belong to the high-speed driving scenario. In the high-speed driving scenario, by executing the above steps S101 to S104, the predicted value of the sub-network output corresponding to the optimal state is obtained, and the predicted value is used as the optimal prediction result in the driving scenario, and then the power output in the high-speed driving scenario is executed.

[0090] For the ramp driving scenario, the main focus is on dealing with the situation where the ramp driving has a large change, focusing on adjusting the vehicle torque demand, focusing on vehicle speed, acceleration, and slope, and performing weighted processing on the current detection data in the ramp driving scenario, which is executed through the following formula:

[0091] T adj (t)=γ1·θ(t)+γ2·v(t)+γ3·T(t)+γ4·a(t)

[0092] Among them, γ1 is the weight corresponding to the slope, γ2 is the weight corresponding to the speed, γ3 is the weight corresponding to the torque, θ(t) is the slope, v(t) is the speed, T(t) is the torque, a(t) is the acceleration, T adj(t) is the torque requirement value under this slope driving scenario.

[0093] By weighting the current detection data in the ramp driving scenario, it is convenient to timely process the power output of the vehicle in the ramp driving scenario. For example, when v(t) and θ(t) meet specific requirements, it is determined to be in the ramp driving scenario. In the ramp driving scenario, by executing the above steps S101 to S104, the predicted value of the sub-network output corresponding to the optimal state is obtained, and the predicted value is used as the optimal prediction result in the ramp driving scenario, and then the power output in the ramp driving scenario is executed.

[0094] In this embodiment, a method for predicting power output data of a vehicle in multiple driving scenarios is provided, which can be used in computer devices, such as mobile phones, tablet computers, desktop computers, portable notebooks, servers, etc. In the above step S101, current detection data of the vehicle in the current driving scenario is obtained, such as Figure 3 As shown, the process includes the following steps:

[0095] Step S301, obtaining data detection results of multiple sensors of the vehicle in the current driving state and current acquisition frequencies of the multiple sensors within a preset time.

[0096] Specifically, the current driving state is a state corresponding to the vehicle being in a driving scenario, and the preset time may be a detection time set for turning on a sensor detection mode in the current driving state.

[0097] In a specific example, the multiple sensors include: a first sensor for detecting vehicle operation data, a second sensor for detecting vehicle road condition data, a third sensor for detecting traffic data, and a fourth sensor for detecting driving environment data.

[0098] Exemplarily, the first type of sensor includes: an acceleration sensor, a vehicle speed sensor, and a generator speed sensor, wherein the acceleration sensor is used to detect vehicle acceleration, and the speed sensor is used to detect vehicle speed, wherein the acceleration sensor transmits data to the electronic control unit via the CAN bus on the vehicle chassis to ensure that the vehicle speed sensor synchronously collects the vehicle speed. The motor speed sensor monitors the speed of the engine and motor, and reflects it to the data monitoring system in real time, and monitors the current driving force and battery power. The data detection results detected by the above-mentioned first type of sensor belong to vehicle operation data. For example, the acceleration Where v(t) is the vehicle speed and t is time. The generator speed sensor collects the engine speed n(t).

[0099] Exemplarily, the second sensor includes: a camera, a laser radar sensor, and a millimeter wave radar sensor, and the slope θ(t) of the road is detected by the second sensor. The second sensor is used to detect vehicle road condition data.

[0100] Exemplarily, the third sensor includes: a sensor for detecting traffic flow and a sensor for detecting traffic light data, and the traffic flow is estimated by the following formula: Among them, N vehicles (t) is the unit road area A at the current moment lane The third type of sensor is used to detect traffic data.

[0101] Exemplarily, the fourth sensor includes: a temperature sensor, through which the current ambient temperature T is detected. e (t). The fourth sensor is used to detect driving environment data.

[0102] The disclosed embodiments obtain data detection results of the vehicle in the current driving state through the above-mentioned various types of multiple sensors, so as to timely know the current operating status of the vehicle in the current driving state, the current road conditions, the current traffic flow conditions, and the current operation status of the driver.

[0103] Step S302, based on the data detection result of the target sensor and the current acquisition frequency of the target sensor, the current data change rate of the target sensor is acquired; wherein the target sensor is any one of a plurality of sensors.

[0104] Specifically, for example, the target sensor is the first sensor among multiple sensors, and the current acquisition frequency of the first sensor is 5 times every 10 minutes. The current data change rate of the first sensor is obtained based on the 5 data detection results of the first sensor. For example, the first sensor is a specific vehicle speed sensor. The vehicle speed of the first detection is 30Km / h, the vehicle speed of the second detection is 31Km / h, the vehicle speed of the third detection is 30Km / h, the vehicle speed of the fourth detection is 33Km / h, and the vehicle speed of the fifth detection is 32Km / h. The change between the first and second times is 1, the change between the second and third times is 1, the change between the third and fourth times is 3, and the change between the fourth and fifth times is 1. The current data change rate of the target sensor is ΔC s (t) = (1 + 1 + 3 + 1) / 4 = 1.5.

[0105] Step S303: Calculate the dynamic adjustment result of the data of the target sensor based on the current acquisition frequency of the target sensor and the current data change rate of the target sensor.

[0106] Step S304: adjusting the current acquisition frequency of the target sensor based on the data dynamic adjustment result.

[0107] Specifically, since the vehicle may encounter sudden environmental changes during driving, such as sudden rainy days, slopes or highways, taking the example of sudden rainy days, if the current environmental humidity is collected at a fixed frequency by the sensor that collects the driving environment, it is impossible to adapt to the changes in the external rainy environment in time, resulting in the drive system being unable to adjust the power output or other key control parameters in time. If the sensor has been collecting data at a fixed frequency when the actual working condition of the vehicle is stable, the collection of a large amount of similar data will easily cause data redundancy and increase the calculation burden of the drive system, thereby affecting the overall efficiency of the drive system. In addition, when multiple sensors work together at the same time, collecting data at a fixed frequency for all sensors will increase system delays and resource consumption. Therefore, it is necessary to flexibly adjust the current acquisition frequency of the target sensor in combination with the actual driving situation.

[0108] In some optional implementations, the dynamic adjustment result of the data of the target sensor is calculated by the following formula;

[0109]

[0110] Among them, f sensor () is the current acquisition frequency of the target sensor, ΔC s (t) is the current data change rate of the target sensor, ΔC th is the preset threshold, f high is the first acquisition frequency, f low The first acquisition frequency may be a high frequency, the second acquisition frequency may be a low frequency, and the target sensor may be any one of a plurality of sensors.

[0111] For example, in heavy rain, the friction coefficient of the road affects the driving performance of the vehicle, so the data priority of the friction coefficient sensor in the fourth sensor for detecting driving environment data will be increased, usually ΔC s (t)≥ΔC th , according to f sensor ()=f high , ensuring data collection at a high frequency. When the air humidity has little impact on the vehicle, the data priority of the humidity sensor in the fourth sensor used to detect the driving environment data will be reduced, ΔC s (t)<ΔC th , according to f sensor = low .

[0112] Ensure data is collected at a low frequency to save system resources and avoid redundant transmission of irrelevant data.

[0113] For another example, when a vehicle is traveling at a constant speed on a flat road, if data is collected at a specific high frequency all the time, data resources will be wasted. At this time, the data priority of the speed sensor in the first sensor used to detect vehicle operation data will be reduced, ΔC s (t)<ΔC th , according to f sensor = low , ensuring that data is collected at a low frequency, thereby saving system resources and avoiding redundant transmission of irrelevant data.

[0114] In this embodiment, a method for predicting power output data of a vehicle in multiple driving scenarios is provided, which can be used in computer devices, such as mobile phones, tablet computers, desktop computers, portable notebooks, servers, etc. After step S301 of obtaining data detection results of multiple sensors of the vehicle in the current driving state and the current acquisition frequencies of the multiple sensors within a preset time, the method can be used to predict power output data of a vehicle in multiple driving scenarios. Figure 4 As shown, the process also includes the following steps:

[0115] Step S401, calculating a multi-sensor collaborative detection result based on the data detection results of the multiple sensors and the weight parameters of the multiple sensors.

[0116] In some optional implementations, the multi-sensor collaborative detection result is calculated by the following formula:

[0117]

[0118] Among them, O sync (t) is the multi-sensor collaborative detection parameter, w i is the weight parameter of the i-th sensor, R i (t) is the detection result of the i-th sensor, and n is the total number of sensors.

[0119] When the data detection result of a certain sensor is in a stable state during the driving process of the vehicle, the data detection result of the sensor can be shared with other sensors, thereby avoiding repeated collection of interrelated data.

[0120] Step S402: based on the data dynamic adjustment result and the multi-sensor collaborative detection result, the data sharing results of the multiple sensors are obtained.

[0121] In a specific example, the above step S402, based on the data dynamic adjustment result and the multi-sensor collaborative detection result, obtains the data sharing results of multiple sensors, including:

[0122] Step d1, based on the data dynamic adjustment result, obtain the current acquisition frequency of the target sensor after dynamic adjustment.

[0123] Specifically, for example, in heavy rain, the friction coefficient of the road affects the driving performance of the vehicle. Therefore, the data priority of the friction coefficient sensor in the fourth sensor for detecting the driving environment data is increased, usually ΔC s (t)≥ΔC th , according to f sensor ()=f high , ensuring that data is collected at a high frequency. At this time, the data of the target sensor is dynamically adjusted to f sensor ()=f high , the current acquisition frequency of the target sensor after dynamic adjustment is f high When the air humidity has little impact on the vehicle, the data priority of the humidity sensor in the fourth sensor that detects the driving environment data will be reduced, ΔC s (t)<ΔC th , according to f sensor = low At this time, the dynamic adjustment result of the target sensor data is f sensor ()=f low .

[0124] Step d2: if the current acquisition frequency of the target sensor after dynamic adjustment remains stable within a preset time, the data detection result of the target sensor after multi-sensor coordination is obtained in a stable state;

[0125] Step d3, using the data detection result of the target sensor after multi-sensor collaboration as the data sharing result of multiple sensors.

[0126] Specifically, for example, when the vehicle is driving uphill, road visual information is obtained through the camera, lidar sensor, and millimeter-wave radar sensor in the second sensor. Since the camera, lidar sensor, and millimeter-wave sensor can all be used to detect road visual information, when the current acquisition frequency of the target sensor after dynamic adjustment of the camera remains stable within the preset time, the data detection results of the camera can be used as the data sharing results of multiple sensors.

[0127] The disclosed embodiment obtains the data dynamic adjustment result of the target sensor, which is conducive to flexibly adjusting the current acquisition frequency of the target sensor, facilitating timely adaptation to the rapidly changing external environment, thereby enabling the drive system to adjust the power output or other key control parameters in a timely manner. At the same time, it can also avoid the situation where the actual working condition of the vehicle is stable, and the data redundancy and increase in the calculation burden of the drive system due to the collection of a large amount of similar data are easily caused, thereby significantly improving the overall efficiency of the drive system. In addition, the disclosed embodiment combines the data dynamic adjustment result, and by calculating the multi-sensor collaborative detection result, it is conducive to realizing collaborative computing and information sharing between multiple sensors, thereby improving the efficiency of the overall system and the effectiveness of data processing, and ultimately enabling the drive system to make efficient intelligent responses in complex environments.

[0128] like Figure 5 As shown, it is a simple block diagram of the vehicle in the embodiment of the present disclosure collecting the current vehicle detection data in real time under multiple driving scenarios. Figure 5 In the current driving state, the vehicle collects data detection results of multiple sensors in real time, mainly including five types of detection data, namely vehicle operation data, vehicle road condition data, traffic data, and driving environment data. These five types of data detection results are pre-processed and optimized. After the optimization process, the data sharing results of multiple sensors are obtained based on the data dynamic adjustment results and multi-sensor collaborative detection results.

[0129] In this embodiment, a method for predicting power output data of a vehicle in multiple driving scenarios is provided, which can be used in computer devices such as mobile phones, tablet computers, desktop computers, portable notebooks, servers, etc. Figure 6 As shown, the process also includes the following steps:

[0130] Step S601: pre-processing the data detection results of various sensors.

[0131] In some optional implementations, preprocessing the data detection results of multiple sensors includes:

[0132] Step b1, preprocessing the data detection results of various sensors.

[0133] In a specific example, the preprocessing method includes an outlier processing method and a missing value filling processing method.

[0134] Specifically, because the vehicle may be affected by environmental interference or hardware errors during driving, unreasonable abnormal values ​​may appear in the sensor data detection results. For example, the temperature sensor may have a sudden change in a stable environment, and such abnormal data needs to be identified and filtered.

[0135] In a specific example, the outlier processing method is to use the mean standard deviation method to identify outliers, and the mean standard deviation method is expressed by the following formula:

[0136]

[0137] When Z i (t) exceeds a predetermined threshold (|Z i (t)|>3), the data is considered as an outlier and is processed.

[0138] In a specific example, the missing value filling processing method is to use linear interpolation to fill in the missing values, and the linear interpolation method is expressed by the following formula:

[0139]

[0140] Among them, X i (t prev ) is the normal data point before the missing value, X i (t next ) is the normal data point after the missing value, t prev is the time point before the missing value, t next is the time point corresponding to the missing value, and t is the current time point.

[0141] Step b2: reprocessing the preprocessed data detection results of the various sensors according to the adaptive Kalman filter algorithm.

[0142] In a specific example, the adaptive Kalman filter algorithm is implemented by the following formula:

[0143]

[0144] K(t)=P(t|t-1)H T (HP(t|t-1)H T +R(t)) -1

[0145]

[0146] in, is the state estimation matrix of the i-th sensor at time t-1, F is the state transfer matrix of the i-th sensor at time t-1, Bu(t) is the control input matrix at time t, K(t) is the Kalman gain matrix at time t, P is the covariance matrix of the i-th sensor at time t-1, H T is the observation matrix, R(t) is the observation noise covariance matrix, is the state estimation matrix of the i-th sensor at time t, X i (t) is the prediction result of the i-th sensor at time t.

[0147] Specifically, in the above formula, the state estimation matrix at the next moment is predicted by the state transfer matrix F and the control input matrix Bu(t), and further based on the error covariance matrix P and the observation matrix H T , observe the noise covariance matrix R(t), calculate the Kalman gain K(t), and use the Kalman gain K(t) to update the current state estimation matrix To make it closer to the actual observation value, in the disclosed embodiment, the Kalman gain is adjusted in real time through the adaptive Kalman filter algorithm to achieve dynamic adjustment of the sensor's data detection results. The error covariance matrix is ​​used to estimate the difference between the current observation value and the predicted value, and to correct the sensor's data detection results based on the difference, so that the updated estimate is closer to the actual data, and finally achieves a data optimization prediction effect. For example, when the sensor cry noise is small or the data is relatively stable, the Kalman gain K(t) will decrease, reducing the current amplitude of the sensor to adapt to the current changes; for example, when the noise is large or the data fluctuates significantly, the Kalman gain K(t) will increase, reducing the current amplitude of the sensor to adapt to the current changes.

[0148] Step S602, obtaining acceleration and velocity from data detection results of various sensors.

[0149] Step S603: Create physical constraints of acceleration and velocity using the following formula.

[0150] v min ≤v(t)≤v max

[0151] a min ≤a(t)≤a max

[0152] Step S604: Based on the physical constraints of acceleration and velocity, a preset optimization algorithm is expressed by the following formula.

[0153]

[0154] Among them, V min is the minimum speed value, V max is the first speed threshold, V max is the second speed threshold, v(t) is the speed at time t, a min is the first acceleration threshold, a max The second acceleration threshold, v(t) is the acceleration at time t;

[0155] Step S605: optimizing the filtered data detection results of the various sensors according to a preset optimization algorithm.

[0156] Specifically, since the speed and acceleration of a vehicle are directly related to its dynamic performance during driving, physical constraints of acceleration and speed are created. The preset optimization algorithm ensures that the filtered data is more consistent with the actual operation of the vehicle, thereby improving the accuracy of the data detection results of multiple sensors.

[0157] In this embodiment, a device for predicting power output data of a vehicle in multiple driving scenarios is also provided, which is used to implement the above-mentioned embodiments and preferred implementation modes, and will not be repeated hereafter. As used below, the term "module" may be a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0158] This embodiment provides a vehicle power output data prediction device in multiple driving scenarios, such as Figure 7 As shown, the device comprises:

[0159] The first acquisition module 701 is used to acquire the current detection data, historical average power demand data, and historical energy recovery time data of the vehicle in the current driving scenario. The current detection data includes: vehicle operation data, vehicle road condition data, traffic data, and driving environment data. The current driving scenario includes a city congestion scenario, a high-speed driving scenario, or a slope driving scenario.

[0160] The data prediction module 702 is used to predict the power output power prediction value corresponding to the vehicle road condition data through the first sub-network, predict the torque demand prediction value corresponding to the traffic data and the driving environment data through the second sub-network combined with the power output power prediction value, and predict the long-term power demand prediction value corresponding to the historical average power demand data and the energy recovery time prediction value corresponding to the historical energy recovery time data through the third sub-network, wherein the first sub-network, the second sub-network and the third sub-network constitute a multi-layer network structure;

[0161] The second acquisition module 703 is used to acquire the sub-network corresponding to the optimal state of the vehicle in the current driving scenario based on the power output power prediction value, the torque demand prediction value, the long-term power demand prediction value, and the energy recovery time prediction value;

[0162] The third acquisition module 704 is used to obtain the predicted value of the sub-network output corresponding to the optimal state, and use the predicted value as the optimal prediction result in the current driving scenario.

[0163] The further functional description of each of the above modules and units is the same as that of the above corresponding embodiments and will not be repeated here.

[0164] The power output data prediction device for the vehicle in multiple driving scenarios in this embodiment is presented in the form of a functional unit, where the unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that executes one or more software or fixed programs, and / or other devices that can provide the above functions.

[0165] An embodiment of the present invention further provides a computer device having the above-mentioned device for predicting power output data of a vehicle in multiple driving scenarios.

[0166] See also Figure 8 , Figure 8 is a schematic diagram of the structure of a computer device provided by an optional embodiment of the present invention, such as Figure 8 As shown, the computer device includes: one or more processors 10, a memory 20, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. Various components are connected to each other using different buses for communication, and can be installed on a common mainboard or installed in other ways as needed. The processor can process the instructions executed in the computer device, including instructions stored in or on the memory to display the graphical information of the GUI on an external input / output device (such as, a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple computer devices can be connected, and each device provides some necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Figure 8 A processor 10 is taken as an example.

[0167] The processor 10 may be a central processing unit, a network processor or a combination thereof. The processor 10 may further include a hardware chip. The hardware chip may be a dedicated integrated circuit, a programmable logic device or a combination thereof. The programmable logic device may be a complex programmable logic device, a field programmable gate array, a general purpose array logic or any combination thereof.

[0168] The memory 20 stores instructions executable by at least one processor 10, so that the at least one processor 10 executes the method shown in the above embodiment.

[0169] The memory 20 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created according to the use of the computer device, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some optional embodiments, the memory 20 may optionally include a memory remotely arranged relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0170] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk or a solid state drive; the memory 20 may also include a combination of the above types of memory.

[0171] The computer device further comprises a communication interface 30 for the computer device to communicate with other devices or a communication network.

[0172] The embodiment of the present invention also provides a computer-readable storage medium. The method according to the embodiment of the present invention can be implemented in hardware, firmware, or can be implemented as a computer code that can be recorded in a storage medium, or can be implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and will be stored in a local storage medium through a network download, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state hard disk, etc.; further, the storage medium can also include a combination of the above types of memories. It can be understood that a computer, a processor, a microprocessor controller, or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor, or hardware, the method shown in the above embodiment is implemented.

[0173] A part of the present invention may be applied as a computer program product, such as a computer program instruction, which, when executed by a computer, can call or provide the method and / or technical solution according to the present invention through the operation of the computer. Those skilled in the art should understand that the existence of the computer program instruction in a computer-readable medium includes, but is not limited to, a source file, an executable file, an installation package file, etc., and accordingly, the way in which the computer program instruction is executed by the computer includes, but is not limited to: the computer directly executes the instruction, or the computer compiles the instruction and then executes the corresponding compiled program, or the computer reads and executes the instruction, or the computer reads and installs the instruction and then executes the corresponding installed program. Here, the computer-readable medium may be any available computer-readable storage medium or communication medium accessible to the computer.

[0174] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A method for predicting power output data of a vehicle in multiple driving scenarios, characterized in that: The method comprises: Obtaining the current detection data, historical average power demand data, and historical energy recovery time data of the vehicle in the current driving scenario, wherein the current detection data includes: vehicle operation data, vehicle road condition data, traffic data, and driving environment data; The first sub-network predicts the power output power prediction value corresponding to the vehicle road condition data, the second sub-network predicts the traffic data and the torque demand prediction value corresponding to the driving environment data in combination with the power output power prediction value, and the third sub-network predicts the long-term power demand prediction value corresponding to the historical average power demand data and the energy recovery time prediction value corresponding to the historical energy recovery time data, respectively, wherein the first sub-network, the second sub-network and the third sub-network form a multi-layer network structure; Based on the power output power prediction value, the torque demand prediction value, the long-term power demand prediction value, and the energy recovery time prediction value, obtaining a subnetwork corresponding to the optimal state of the vehicle in the current driving scenario; Obtain a predicted value of the subnetwork output corresponding to the optimal state, and use the predicted value as the optimal prediction result in the current driving scenario.

2. The method according to claim 1, characterized in that The predicting, by using the first sub-network, a first load prediction value corresponding to the vehicle road condition data includes: Obtaining speed, acceleration, and engine speed from the vehicle road condition data; Based on the speed, the acceleration and the engine speed, calculating the power output power prediction value corresponding to the vehicle road condition data predicted by the first sub-network, wherein the power output power prediction value is calculated by the following formula; P pred (t)=f conv (v(t),a(t),n(t)) Among them, P pred (t) is the predicted value of the power output, f conv (v(t) is the convolution operation factor, v(t) is the speed, a(t) is the acceleration, and n(t) is the engine speed.

3. The method according to claim 1, characterized in that The predicting of the torque demand prediction value corresponding to the vehicle road condition data, the traffic data, and the driving environment data by combining the power output power prediction value through the second sub-network includes: Acquire the slope from the vehicle road condition data, acquire the traffic flow from the traffic data, and acquire the temperature from the driving environment data; Based on the power output power prediction value, the slope, the traffic flow, and the temperature, the torque demand prediction value corresponding to the traffic data and the driving environment data predicted by the second sub-network is calculated, wherein the torque demand prediction value is calculated by the following formula; B pred (t)=f merge (P pred (t),θ(t),T(t),T f (t)) Among them, B pred (t) is the torque demand prediction value, f merge is the feature fusion factor, P pred (t) is the predicted value of the power output, θ(t) is the slope, T(t) is the temperature, and Tf(t) is the traffic flow.

4. The method according to claim 1, characterized in that: The method of respectively predicting the long-term power demand prediction value corresponding to the historical average power demand data and the energy recovery time prediction value corresponding to the historical energy recovery time data by the third sub-network includes: Based on the historical average power demand data and the historical energy recovery time data, the long-term power demand forecast value and the energy recovery time forecast value are calculated, wherein the long-term power demand forecast value and the energy recovery time forecast value are calculated by the following formula; P long (t),E long (t)=f LSTM (P avg (t),E rec (t)) Among them, P long (t) is the long-term power demand forecast value, E pred (t) is the predicted value of the energy recovery time, f LSTM is the calculation factor of the third sub-network, P avg (t) is the historical average power demand data, E rec (t) is the historical energy recovery time data.

5. The method according to claim 1, characterized in that Based on the power output power prediction value, the torque demand prediction value, the long-term power demand prediction value, and the energy recovery time prediction value, the subnetwork corresponding to the optimal state of the vehicle in the current driving scenario is obtained, which is executed by the following formula: Wherein, S(t) is the state feature vector corresponding to the current detection data of the vehicle in the current driving scene monitored in real time so that the current state of the vehicle reaches the optimal state, and f i (X(t), Y(t)) is the output accuracy of the i-th subnetwork corresponding to the optimal state in the current driving scenario, X(t) is the input feature of the i-th subnetwork, and Y(t) is the predicted value output by the i-th subnetwork.

6. The method according to claim 1, characterized in that The current driving scenario includes a city congestion scenario, a high-speed driving scenario, or a ramp driving scenario, and further includes: The current detection data in the urban congestion scenario, the high-speed driving scenario, or the slope driving scenario is weighted.

7. A vehicle power output data prediction device in multiple driving scenarios, characterized in that: The device comprises: A first acquisition module is used to acquire current detection data, historical average power demand data, and historical energy recovery time data of the vehicle in the current driving scenario, wherein the current detection data includes: vehicle operation data, vehicle road condition data, traffic data, and driving environment data, and the current driving scenario includes a city congestion scenario, a high-speed driving scenario, or a slope driving scenario; A data prediction module, used to predict the power output power prediction value corresponding to the vehicle road condition data through a first sub-network, predict the traffic data and the torque demand prediction value corresponding to the driving environment data through a second sub-network combined with the power output power prediction value, and respectively predict the long-term power demand prediction value corresponding to the historical average power demand data and the energy recovery time prediction value corresponding to the historical energy recovery time data through a third sub-network, wherein the first sub-network, the second sub-network and the third sub-network constitute a multi-layer network structure; A second acquisition module is used to acquire a subnetwork corresponding to the optimal state of the vehicle in the current driving scenario based on the power output power prediction value, the torque demand prediction value, the long-term power demand prediction value, and the energy recovery time prediction value; The third acquisition module is used to obtain the predicted value of the sub-network output corresponding to the optimal state, and use the predicted value as the optimal prediction result in the current driving scene.

8. A computer device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the method for predicting power output data of a vehicle in multiple driving scenarios according to any one of claims 1 to 6 by executing the computer instructions.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the method for predicting power output data of a vehicle in multiple driving scenarios according to any one of claims 1 to 6.

10. A computer program product, characterized in that The method comprises computer instructions for causing a computer to execute the method for predicting power output data of a vehicle in multiple driving scenarios according to any one of claims 1 to 6.

Images