Power consumption control method, device and equipment of automatic driving system and storage medium

Abstract

The invention discloses a power consumption control method and device of an automatic driving system, equipment and a storage medium. The method comprises the following steps: acquiring road information and vehicle information of a target vehicle in a current driving environment; and road scene features are extracted from the road information, vehicle attitude control features are extracted from the vehicle information, and based on the extracted features, a proportion prediction model obtained by training is used to predict the demand proportion of the target vehicle to the sensor in the current road scene. And searching a matched power consumption control state level from a preset power consumption control state grading library according to the predicted demand ratio. And determining a target driving mode and a target sensor based on the driving mode and the operable sensor identifier in the searched power consumption control state level, controlling the target vehicle to operate the target sensor, and performing automatic driving control in the target driving mode. Therefore, the prediction accuracy of the demand ratio is improved, hierarchical management of energy consumption is realized, and the power consumption of the whole vehicle is reduced.

Description

Technical Field

[0001] The present application relates to the field of autonomous driving technology, specifically to technical fields such as power consumption control and sensor management, and in particular to a power consumption control method, device, equipment and storage medium for an autonomous driving system. Background Art

[0002] Autonomous driving uses advanced communication, computer, network and control technologies, and relies on the architecture of multi-sensor fusion and high-computing power platform to achieve real-time and continuous control of vehicles. This technology is a redundant design for autonomous driving in all scenarios, so as to meet extremely complex and extreme scenarios of autonomous driving through multi-sensor data fusion and high-computing power data processing platform.

[0003] However, this redundant design, that is, the design that requires multiple sensors and high-computing data processing platforms to assist in data processing regardless of the scenario, directly leads to higher energy consumption required for vehicle operation and additional energy consumption caused by heat dissipation required for running too many devices or components, increasing the vehicle's power consumption during the autonomous driving process. Summary of the invention

[0004] The present application provides a power consumption control method, device, equipment and storage medium for an autonomous driving system to solve the problem of high power consumption caused by the operation of redundant multi-sensors and high-computing power data processing platforms in order to adapt to a variety of complex autonomous driving scenarios in existing autonomous driving systems.

[0005] The technical solution is as follows:

[0006] In a first aspect, a method for controlling power consumption of an autonomous driving system is provided, comprising:

[0007] Obtain road information and vehicle information of the target vehicle in the current driving environment;

[0008] Extracting road scene features from the road information and extracting vehicle posture control features from the vehicle information, and based on the road scene features and the vehicle posture control features, using the trained share prediction model, predicting the sensor demand share of the target vehicle in the current road scene;

[0009] According to the predicted demand ratio, searching for a matching power consumption control state level from a preset power consumption control state classification library; wherein the preset power consumption control state classification library stores a plurality of power consumption control state levels, each power consumption control state level being configured with a driving mode, a demand ratio, and an operable sensor identifier;

[0010] Based on the driving mode and the operable sensor identifiers in the found power consumption control state level, a target driving mode and a target sensor are determined respectively, the target vehicle is controlled to operate the target sensor, and automatic driving control is performed in the target driving mode.

[0011] In a possible implementation, based on the road scene features and the vehicle posture control features, the trained share prediction model is used to predict the sensor demand share of the target vehicle in the current road scene, specifically including:

[0012] The road scene features and the vehicle posture control features are respectively input into the trained proportion prediction model to predict the proportion of sensor demand of the target vehicle in the current road scene; wherein the trained proportion prediction model is a multimodal prediction model.

[0013] In a possible implementation, the proportion prediction model is trained in the following way:

[0014] Obtaining historical data of autonomous driving of sample vehicles of the same model as the target vehicle when running at full power consumption in various road scenarios;

[0015] A sample data set for training is selected from the autonomous driving historical data; wherein the sample data set includes a plurality of sample data and a sample label corresponding to each sample data, and each sample data includes at least: a road scene feature and a vehicle posture control feature; and the sample label is an operating ratio of a plurality of sensors on the sample vehicle;

[0016] Inputting a plurality of sample data consisting of road scene features and vehicle posture control features and sample labels in the sample data set into a preset neural network model for training;

[0017] Based on the optimal control strategy, the model parameters are repeatedly adjusted to obtain a converged share prediction model.

[0018] In a possible implementation, the method further includes:

[0019] Periodically detect the road information of the target vehicle in the current driving environment;

[0020] If a change in road information is detected, the latest road information and vehicle information of the target vehicle in the current driving environment are re-acquired; and based on the latest road information and vehicle information, the road scene feature and vehicle posture control feature are respectively extracted, and the trained proportion prediction model is used to predict the latest demand proportion of the target vehicle for the sensor in the current road scene;

[0021] According to the predicted latest demand ratio, searching for the matching latest power consumption control state level from a preset power consumption control state classification library;

[0022] The target vehicle is controlled to switch from a current power consumption control state level to a latest power consumption control state level.

[0023] In a possible implementation, the method further includes:

[0024] Periodically receive vehicle control data and autonomous driving control data and analyze them;

[0025] Evaluate whether the power consumption control state level of the target vehicle in the current driving environment matches according to the analysis result;

[0026] If they do not match, the power consumption control state level is exited, triggering entry into the direct control mode between the vehicle control system and the autonomous driving system.

[0027] In a second aspect, a power consumption control device of an autonomous driving system is provided, comprising:

[0028] An acquisition module is used to acquire road information and vehicle information of a target vehicle in a current driving environment;

[0029] A prediction module, used to extract road scene features from the road information, and extract vehicle posture control features from the vehicle information, and based on the road scene features and the vehicle posture control features, use the trained proportion prediction model to predict the proportion of sensor demand of the target vehicle in the current road scene;

[0030] A search module, configured to search for a matching power consumption control state level from a preset power consumption control state classification library according to the predicted demand proportion; wherein the preset power consumption control state classification library stores a plurality of power consumption control state levels, each power consumption control state level being configured with a driving mode, a demand proportion, and an operable sensor identifier;

[0031] The control module is used to determine the target driving mode and the target sensor respectively based on the driving mode and the operable sensor identifier in the power consumption control state level found, control the target vehicle to operate the target sensor, and perform automatic driving control in the target driving mode.

[0032] In a third aspect, a computer-readable storage medium is provided, wherein at least one instruction is stored in the storage medium, and the at least one instruction is loaded and executed by a processor to implement the method of the above-mentioned aspect and any possible implementation manner.

[0033] In a fourth aspect, an electronic device is provided, including:

[0034] at least one processor; and

[0035] a memory communicatively connected to the at least one processor; wherein,

[0036] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method of any possible implementation manner and the aspects described above.

[0037] According to a fifth aspect, a computer program product is provided, comprising a computer program, which, when executed by a processor, implements the above-mentioned aspects and any possible implementation method.

[0038] In a sixth aspect, an autonomous driving vehicle is provided, comprising the electronic device as described above.

[0039] The beneficial effects of the technical solution provided by this application include at least:

[0040] It can be seen from the above technical solution that the embodiment of the present application obtains the road information and vehicle information of the target vehicle in the current driving environment; extracts the road scene features from the road information, and extracts the vehicle posture control features from the vehicle information, and based on the road scene features and the vehicle posture control features, uses the trained proportion prediction model to predict the proportion of sensor demand of the target vehicle in the current road scene. Then, according to the predicted demand proportion, the matching power consumption control state level is searched from the preset power consumption control state classification library. After that, based on the driving mode and the operable sensor identifier in the found power consumption control state level, the target driving mode and the target sensor are determined respectively, the target vehicle is controlled to operate the target sensor, and the automatic driving control is performed under the target driving mode. Thereby, the prediction accuracy of the demand proportion of the sensor is improved, and under the condition of ensuring the safety of the automatic driving of the automatic driving vehicle, the hierarchical management of energy consumption is realized by controlling the number of sensors running and the open and close states in different working modes to reduce energy consumption.

[0041] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0043] Figure 1 It is a schematic diagram of the steps of the power consumption control method of the autonomous driving system provided in an embodiment of the present application.

[0044] Figure 2 It is a schematic diagram of the power consumption control flow of the autonomous driving system provided in an embodiment of the present application.

[0045] Figure 3 This is a schematic diagram of the structure of a power consumption control device for an autonomous driving system provided by an embodiment of the present application.

[0046] Figure 4 This is a structural block diagram of a power consumption control system of an autonomous driving system provided in yet another embodiment of the present application.

[0047] Figure 5 It is a block diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0048] The following is a description of exemplary embodiments of the present application in conjunction with the accompanying drawings, including various details of the embodiments of the present application to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present application. Similarly, for the sake of clarity and conciseness, the description of well-known functions and structures is omitted in the following description.

[0049] Obviously, the described embodiments are only part of the embodiments of the present application, but not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.

[0050] It should be noted that the terminal devices involved in the embodiments of the present application may include but are not limited to mobile phones, personal digital assistants (PDAs), wireless handheld devices, tablet computers and other smart devices; display devices may include but are not limited to personal computers, televisions and other devices with display functions.

[0051] In addition, the term "and / or" in this article is only a description of the association relationship between the associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0052] The current technology routes for the development of autonomous driving mostly rely on the architecture of multi-sensor fusion and high-computing power platforms. The essence of this is to make redundant designs for autonomous driving in all scenarios, so as to meet extremely complex and extreme scenarios of autonomous driving through multi-sensor data fusion and high-computing power data processing platforms. For example, in some scenarios, only data from a certain sensor is needed, but all sensors are in working state, which leads to high energy consumption for data processing and additional energy consumption caused by the need to dissipate heat for the computing platform, which greatly increases the power consumption of the vehicle.

[0053] In view of this, the embodiment of the present application provides a power consumption control scheme for an automatic driving system, and its main inventive concept is to obtain the road information and vehicle information of the target vehicle in the current driving environment; extract the road scene features from the road information, and extract the vehicle posture control features from the vehicle information, and based on the road scene features and the vehicle posture control features, use the trained proportion prediction model to predict the demand proportion of the target vehicle for the sensor in the current road scene. Then, according to the predicted demand proportion, the matching power consumption control state level is searched from the preset power consumption control state classification library. After that, based on the driving mode and the operable sensor identifier in the found power consumption control state level, the target driving mode and the target sensor are determined respectively, the target vehicle is controlled to operate the target sensor, and the automatic driving control is performed in the target driving mode. Thus, under the condition of ensuring the safety of the automatic driving of the automatic driving vehicle, by controlling the operating quantity and opening and closing state of sensors, computing power platforms, auxiliary heat dissipation devices, etc. in different working modes, the hierarchical management of energy consumption is realized to reduce energy consumption.

[0054] Reference Figure 1 As shown, it is a schematic diagram of the steps of the power consumption control method of the automatic driving system provided in the embodiment of the present application. The executor of the power consumption control method is the power consumption control device of the automatic driving system. In the present application, the power consumption control device of the automatic driving system can be a software module with computing and data processing capabilities, or it can be an electronic device integrated with a similar software module. In fact, the device is particularly effective in functional low-speed automatic driving vehicles, such as automatic driving shuttles, logistics vehicles, and sweepers. It can determine the minimum working boundary required for the optimal automatic driving according to different working conditions, and interact with the automatic driving system according to the boundary, enter different working modes, and adopt different automatic driving strategies to achieve the purpose of controlling energy consumption.

[0055] like Figure 1 As shown, the power consumption control method of the automatic driving system may specifically include the following steps:

[0056] Step 102: Obtain road information and vehicle information of the target vehicle in the current driving environment.

[0057] In this application, the target vehicle may be a shuttle vehicle, logistics vehicle, delivery vehicle, road sweeper, etc. with automatic driving function; the road information in the current driving environment may include road signs, road layout, traffic signals, lane lines and other information. The vehicle information may include the speed, body posture, driving direction, braking status and other driving-related information of the target vehicle. It should be understood that, if conditions permit, the vehicle information may also include driving information of other vehicles in the current driving environment and obstacle information.

[0058] Step 104: extracting road scene features from the road information and extracting vehicle posture control features from the vehicle information, and based on the road scene features and the vehicle posture control features, using the trained share prediction model, predicting the sensor demand share of the target vehicle in the current road scene.

[0059] Afterwards, the road scene in the current driving environment is determined based on the acquired road information, and the road scene features are extracted. For example, highway scenes, expressway scenes, urban road scenes, rural road scenes, tunnel scenes, viaduct scenes, cross-sea bridge scenes, etc. Each different road scene has different road scene features, and the road scene features can be used as feature vectors of the road scene. At the same time, the vehicle posture can be screened out from the vehicle information, and the vehicle posture control features used to characterize the vehicle posture vector can be extracted based on the vehicle posture. The vehicle posture control features can be used as feature vectors of the vehicle posture. The extracted road scene features and vehicle posture control features are sent to the trained proportion prediction model, and the proportion of sensor demand of the target vehicle in the current road scene is predicted through model processing.

[0060] In this application scheme, the demand proportion of sensors can be predicted in the following two ways.

[0061] The first method is to fuse the road scene features and the vehicle posture control features; input the fused features into the trained proportion prediction model to predict the proportion of sensor requirements of the target vehicle in the current road scene.

[0062] The second method: the road scene features and the vehicle posture control features are respectively input into the trained proportion prediction model to predict the proportion of sensor requirements of the target vehicle in the current road scene; wherein the trained proportion prediction model is a multimodal prediction model.

[0063] The first method mentioned above is to first extract two features, then fuse the two different features, and then input the fused features into the trained proportion prediction model for prediction. Accordingly, in this method, the proportion prediction model is also repeatedly trained through the fused features during the model training stage.

[0064] The second method mentioned above is not to fuse the two features, but to input the two features into the trained proportion prediction model at the same time for prediction. In this method, the proportion prediction model is repeatedly trained based on the two features during the model training stage.

[0065] Regardless of which method is used for prediction, the accuracy and processing efficiency of determining the proportion of sensor demand can be improved based on the learning algorithm, which in turn facilitates the subsequent control of the appropriate number of sensors to work and saves the power consumption of the entire vehicle.

[0066] In the present application, the second method can be selected for prediction. Accordingly, the corresponding proportion prediction model of this method can be trained through the following steps:

[0067] The first step is to obtain historical data of autonomous driving of sample vehicles of the same model as the target vehicle when running at full power consumption in various road scenarios.

[0068] First, the model of the target vehicle to be sampled needs to be determined, and one or more sample vehicles for testing are selected based on the determined model. Then, the sample vehicles are controlled to run in different road scenarios, and at this time, the sample vehicles need to be controlled to run at full power consumption. After that, the historical data of autonomous driving of the sample vehicles when running at full power consumption in various road scenarios is collected.

[0069] The autonomous driving historical data may include: vehicle information such as the speed, driving direction, braking status, steering, and body posture of the sample vehicle, as well as road information such as the road layout, road signs, traffic signals, and lane lines in the road scene where the sample passed. In addition, it also includes the number of sensors or sensor identifications that actually work when the sample vehicle is running at full power.

[0070] The second step is to select a sample data set for training from the autonomous driving historical data; wherein the sample data set includes multiple sample data and a sample label corresponding to each sample data, and each sample data includes at least: road scene features and vehicle posture control features; the sample label is the operating ratio of multiple sensors on the sample vehicle.

[0071] It should be understood that each sample vehicle can obtain a set of autonomous driving historical data by driving in each road scene. In this way, a sufficient amount of autonomous driving historical data can be obtained by driving multiple sample vehicles in multiple road scenes. The road scene features and vehicle posture control features are selected from each set of autonomous driving historical data as sample data for training, and the sensor operation ratio is determined as the sample label of the sample data according to the number of sensors or sensor identifications actually working when the sample vehicle is running at full power in the autonomous driving historical data.

[0072] In the third step, multiple sample data consisting of road scene features and vehicle posture control features and sample labels in the sample data set are input into a preset neural network model for training.

[0073] After obtaining a large amount of sample data (road scene features and vehicle posture control features) and sample labels (operation ratio of sensors), they are input into the preset neural network model for model training. The preset neural network model here can be a feedback neural network model such as a long short-term memory network model LSTM, a recurrent neural network model RNN, or a radial neural network model RBF, a convolutional neural network model CNN, or other feedforward neural network models. This application does not limit the type of preset neural network model, and preferably LSTM can be used so that a better model effect can be obtained.

[0074] The fourth step is to repeatedly adjust the model parameters based on the optimal control strategy to obtain a converged share prediction model.

[0075] After repeatedly adjusting and learning the preset neural network model based on the optimal control strategy, a converged share prediction model can be obtained. The share prediction model can accurately predict the operating share of the target vehicle's sensors and improve the processing efficiency of the target vehicle.

[0076] Step 106: According to the predicted demand ratio, search for a matching power consumption control state level from a preset power consumption control state classification library; wherein the preset power consumption control state classification library stores multiple power consumption control state levels, and each power consumption control state level is configured with a driving mode, a demand ratio, and an operable sensor identifier.

[0077] In the present application scheme, a power consumption control state classification library can be created in advance, and the power consumption control state classification library can be stored on the Internet of Vehicles platform or other cloud devices, or stored locally in the target vehicle.

[0078] The power consumption control state classification library stores multiple power consumption control state levels, and each power consumption control state level is configured with a driving mode, a demand ratio, and an operable sensor identifier. Among them, the driving modes involved in this application may include: automatic driving mode, remote takeover mode, and manual control mode. The target vehicle can switch between different driving modes according to driving needs to adapt to the current road scene to perform safe and reliable driving operations.

[0079] The specific number of levels can be set according to the proportion of sensor demand or power consumption. For example, if there are 20 corresponding sensors on the whole vehicle, they can be roughly divided into 10 levels, namely 10%, 20%, 30%, 80%, 90%, 100%. For another example, according to the proportion of computing power and sensor power consumption, it can be roughly divided into 4 levels, namely 0%, 20%, 60%, 100% of full power consumption. Each level includes driving mode, demand proportion and operable sensor identification.

[0080] In this application scheme, it can be divided into 4 levels according to the power consumption ratio, that is, there are 4 levels configured in the power consumption control state level library:

[0081] Power consumption control state level 1: When the target vehicle is in remote takeover mode and manual operation mode, the computing platform and sensors of the autonomous driving system are in power-on standby state. In this mode, the power consumption is the lowest (0% of full power consumption).

[0082] Power consumption control state level 2: The target vehicle is in autonomous driving mode, and the autonomous driving system computing unit platform is operating in a calibrated low power consumption mode (20% of full power consumption). Only the front binocular camera and front millimeter-wave radar are working, and other sensors are in standby sleep mode, realizing the autonomous driving capabilities of automatic following and autonomous cruising.

[0083] Power consumption control state level 3: The target vehicle is in the autonomous driving mode. The computing unit of the autonomous driving system operates at a calibrated medium power consumption (60% of the full power consumption). Only the front dual laser radars and the front monocular camera are working, and other sensors are in standby sleep mode, realizing the autonomous driving capabilities of automatic following and autonomous cruising, as well as the obstacle avoidance and lane change functions.

[0084] Power consumption control state level 4: The vehicle in front of the target is in autonomous driving mode, the autonomous driving system computing unit is operating at the calibrated full power consumption, all sensors are in working state, and all high-level autonomous driving functions are realized.

[0085] It should be understood that in the present application, the operating ratio of the sensor can be equal to the power consumption ratio of the sensor and the computing platform, that is, the operating ratio of the sensor is approximately equal to the full power consumption ratio mentioned above.

[0086] In this way, after predicting the operating proportion of the sensor of the target vehicle in the current road scene according to step 104, the power consumption control state level corresponding to the predicted operating proportion of the sensor can be matched from the power consumption control state level library. For example, if the predicted operating proportion of the sensor is 20%, then the power consumption control state level 2 can be selected. For another example, the predicted operating proportion of the sensor is 30%, but there is no such operating proportion value in the power consumption control state level library. At this time, the level closest to the operating proportion value 30% can be selected. 30% is close to 20%, but far from 60%, so the power consumption control state level that best matches the power consumption control state level 2 is selected. For another example, the predicted operating proportion of the sensor is 40%, which is the same as the closeness of 20% and 60%. At this time, in order to ensure the reliable execution of autonomous driving, the power consumption control state level 3 can be selected. In this way, as many sensors as possible can be turned on to ensure the normal execution of autonomous driving, and power consumption can be saved to a certain extent.

[0087] Step 108: Based on the driving mode and the operable sensor identifiers in the found power consumption control state level, determine the target driving mode and the target sensor respectively, control the target vehicle to operate the target sensor, and perform automatic driving control in the target driving mode.

[0088] Since each power consumption control state level has a corresponding driving mode and sensor identification, after determining the matching power consumption control state level, the target driving mode can be determined according to the driving mode under the power consumption control state level, and the target sensor can be determined according to the sensor identification. Thereafter, the target vehicle is controlled to turn on and operate the target sensor, and automatic driving control is performed under the target driving mode.

[0089] Optionally, in the present application, after the above-mentioned scheme is adopted to control the target vehicle to operate the target sensor and perform automatic driving control in the target driving mode, since the road scene is not fixed and may change over time, that is, it is currently on the expressway, and the next moment it may exit the expressway and enter the urban road. Therefore, the road information of the target vehicle in the current driving environment can also be periodically detected; and when a change in the road information is detected, the above steps 102-108 are repeated.

[0090] In other words, if a change in road information is detected, the latest road information and vehicle information of the target vehicle in the current driving environment are re-acquired; and based on the latest road information and vehicle information, the road scene characteristics and vehicle posture control characteristics are respectively extracted, and the trained proportion prediction model is used to predict the latest demand proportion of the target vehicle for sensors in the current road scene; according to the predicted latest demand proportion, the matching latest power consumption control state level is searched from the preset power consumption control state classification library; the target vehicle is controlled to switch from the current power consumption control state level to the latest power consumption control state level. In this way, it can be ensured that the target vehicle can adjust the power consumption control state level in time according to the changes in the road scene, so that the target vehicle always runs a reasonable number of sensors to perform automatic driving control, saving the power consumption of the whole vehicle.

[0091] Furthermore, in the present application scheme, the whole vehicle control data and the automatic driving control data can also be received periodically and analyzed; based on the analysis results, it is evaluated whether the power consumption control state level of the target vehicle in the current driving environment matches; if it does not match, the power consumption control state level is exited, triggering the entry into the direct control mode between the whole vehicle control system and the automatic driving system. This method can evaluate whether the current controlled state of the target vehicle and the power consumption control state level of the automatic driving are reasonable, and force the exit of the power consumption control of the automatic driving when necessary, realize the direct connection between the whole vehicle control system and the automatic driving system, and control the automatic driving of the target vehicle without the power consumption management control authority.

[0092] It should be understood that the sensor involved in this application may be an image acquisition device or a temperature acquisition device or other sensor device installed or integrated on the target vehicle, such as a camera, a laser radar, a temperature sensor, a millimeter wave radar, etc. This application is not limited to this.

[0093] Reference Figure 2 The figure is a schematic diagram of the power consumption control process of the autonomous driving system provided in an embodiment of the present application.

[0094] Step 202: Collect an autonomous driving data set for the same vehicle model under full power consumption and multi-road scenarios.

[0095] Since different vehicle models are equipped with different types and quantities of sensors, when implementing the power consumption control solution, sampling is performed on autonomous driving vehicles of the same model.

[0096] Step 204: Screen the data and determine the training data.

[0097] Filter the sample data required for this model training from the collected autonomous driving dataset.

[0098] Step 206: Extract the proportion of the vehicle's demand for each sensor data in the autonomous driving system.

[0099] Step 208: extract road scene classification and mark it.

[0100] Step 206 and step 208 are mainly used to determine the feature vectors and sample labels involved in model training from the sample data. Specifically, the demand ratio of sensors is extracted as the sample label, and the road scene is extracted as one of the feature vectors. Subsequently, the vehicle posture can be extracted as another feature vector through step 212.

[0101] Step 210: Temporally couple the road scene with the sensor demand ratio.

[0102] Step 212: Input the vehicle posture control historical data set.

[0103] This step is mainly used to extract the feature vector corresponding to the vehicle posture from the vehicle posture control history data set, and input it into the model for training together with the feature vector of the above road scene.

[0104] Step 214: A prediction model is obtained by performing model training using a long short-term memory network (LSTM) and an optimal control strategy.

[0105] Step 216: Obtain the vehicle posture and road scene in the current driving scene.

[0106] In this step, the feature vector to be predicted in the current driving scenario can be extracted from the acquired vehicle posture and road scene.

[0107] Step 218: Input the prediction model to perform prediction and obtain the demand ratio of sensors in the current scenario.

[0108] Step 220: Search the power consumption control state level library for a state level that matches the demand ratio of the sensor.

[0109] Step 222: Perform autonomous driving using the driving mode and sensor identifier corresponding to the matched state level.

[0110] Therefore, under the condition of ensuring the safety of autonomous driving vehicles, by controlling the number of sensors, computing platforms, auxiliary heat dissipation devices, etc. in different working modes and the on and off status, energy consumption can be managed in a hierarchical manner to reduce energy consumption. In addition, when determining the number of sensors in operation, it is predicted by the LSTM model trained by the machine learning algorithm, which significantly improves the accuracy of the estimation of the number of sensors in operation, thereby improving the safety of autonomous driving and saving energy.

[0111] It should be noted that, for the aforementioned method embodiments, for the sake of simplicity, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the present application is not limited by the described order of actions, because according to the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present application.

[0112] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0113] Figure 3 FIG. 1 shows a structural block diagram of a power consumption control device for an automatic driving system provided by an embodiment of the present application. Figure 3 As shown. The power consumption control device 300 of the automatic driving system of this embodiment may include an acquisition module 301, a prediction module 302, a search module 303 and a control module 304. Among them, the acquisition module 301 is used to obtain the road information and vehicle information of the target vehicle in the current driving environment; the prediction module 302 is used to extract the road scene features from the road information, and extract the vehicle posture control features from the vehicle information, and based on the road scene features and the vehicle posture control features, use the trained proportion prediction model to predict the demand proportion of the target vehicle for the sensor in the current road scene; the search module 303 is used to search for a matching power consumption control state level from a preset power consumption control state classification library according to the predicted demand proportion; wherein the preset power consumption control state classification library stores multiple power consumption control state levels, and each power consumption control state level is configured with a driving mode, a demand proportion and an operable sensor identifier; the control module 304 is used to determine the target driving mode and the target sensor respectively based on the driving mode and the operable sensor identifier in the found power consumption control state level, control the target vehicle to operate the target sensor, and perform automatic driving control in the target driving mode.

[0114] It should be noted that part or all of the power consumption control device of the autonomous driving system of this embodiment may be an application located in the local terminal, or may also be a functional unit such as a plug-in or software development kit (SDK) set in the application located in the local terminal, or may also be a processing engine located in a network-side server, or may also be a distributed system located on the network side, for example, a processing engine or distributed system in an autonomous driving platform on the network side, etc. This embodiment does not specifically limit this.

[0115] It is understandable that the application may be a local program (nativeApp) installed on the local terminal, or may be a webpage program (webApp) of a browser on the local terminal, which is not limited in this embodiment.

[0116] Optionally, in a possible implementation of this embodiment, the acquisition module 301 can be specifically used to input the road scene features and the vehicle posture control features into the trained proportion prediction model respectively, to predict the proportion of sensor demand of the target vehicle in the current road scene; wherein, the trained proportion prediction model is a multimodal prediction model.

[0117] Optionally, in a possible implementation of this embodiment, the proportion prediction model is trained in the following manner:

[0118] Obtain historical data of autonomous driving of sample vehicles of the same model as the target vehicle when running at full power consumption in various road scenarios; filter a sample data set for training from the historical data of autonomous driving; wherein the sample data set includes multiple sample data and a sample label corresponding to each sample data, and each sample data includes at least: a road scene feature and a vehicle posture control feature; the sample label is the operating proportion of multiple sensors on the sample vehicle; input the multiple sample data and sample labels consisting of the road scene feature and the vehicle posture control feature in the sample data set into a preset neural network model for training; repeatedly adjust the model parameters based on the optimal control strategy to obtain a converged proportion prediction model.

[0119] Optionally, in a possible implementation of this embodiment, the power consumption control device of the autonomous driving system further includes: a detection module and a switching module; wherein,

[0120] The detection module may be specifically used to periodically detect road information of the target vehicle in the current driving environment;

[0121] The acquisition module 301 is further used to reacquire the latest road information and vehicle information of the target vehicle in the current driving environment if a change in the road information is detected; and the prediction module 302 is further used to extract the road scene feature and vehicle posture control feature based on the latest road information and vehicle information, and use the trained proportion prediction model to predict the latest demand proportion of the target vehicle for the sensor in the current road scene;

[0122] The search module 303 is further used to search for a matching latest power consumption control state level from a preset power consumption control state classification library according to the predicted latest demand proportion;

[0123] The switching module is specifically used to control the target vehicle to switch from the current power consumption control state level to the latest power consumption control state level.

[0124] Optionally, in a possible implementation of this embodiment, the power consumption control device of the autonomous driving system further includes: a parsing module, an evaluation module and an exit module; wherein,

[0125] The analysis module is specifically used to periodically receive the vehicle control data and the automatic driving control data, and perform analysis;

[0126] The evaluation module is specifically used to evaluate whether the power consumption control state level of the target vehicle in the current driving environment matches according to the analysis result;

[0127] The exit module is specifically used to exit the power consumption control state level if there is no match, triggering entry into the direct control mode between the vehicle control system and the automatic driving system.

[0128] In this embodiment, the road information and vehicle information of the target vehicle in the current driving environment can be obtained; the road scene features are extracted from the road information, and the vehicle posture control features are extracted from the vehicle information, and based on the road scene features and the vehicle posture control features, the trained proportion prediction model is used to predict the proportion of sensor demand of the target vehicle in the current road scene. Then, according to the predicted demand proportion, the matching power consumption control state level is searched from the preset power consumption control state classification library. After that, based on the driving mode and the operable sensor identifier in the found power consumption control state level, the target driving mode and the target sensor are determined respectively, the target vehicle is controlled to operate the target sensor, and the automatic driving control is performed in the target driving mode. Thus, under the condition of ensuring the safety of the automatic driving of the automatic driving vehicle, by controlling the operating quantity and opening and closing state of sensors, computing power platforms, auxiliary heat dissipation devices, etc. in different working modes, the hierarchical management of energy consumption is realized to reduce energy consumption.

[0129] Reference Figure 4 4 is a schematic diagram of the structure of the power consumption control system of the autonomous driving system provided in the embodiment of the present application. The power consumption control system may include: an autonomous driving power consumption control device 401, an autonomous driving control system 402, a vehicle control system 403, a vehicle cloud control platform 404 and an autonomous driving sensor system 405.

[0130] The automatic driving power consumption control device 401 can be divided into the following functional modules according to different functions: state detection module, signal transmission module, scene analysis module, mode switching module and safety redundancy module; wherein the state detection module is used to detect the current driving mode of the whole vehicle, and detect the working state and fault state of the sensor included in the automatic driving control system; wherein the driving module includes: automatic driving mode, remote takeover mode, and manual operation mode. The signal transmission module is used to complete the interactive transmission of information with the whole vehicle control system and the automatic driving control system. The scene analysis module is used to obtain a relatively reliable automatic driving classification state based on the state prediction model based on the neural network algorithm according to the current road environment, sensor data, automatic driving state, vehicle historical data and other data. The mode switching module is used to switch different mode working states according to the automatic driving classification state obtained by the scene analysis module, and switch between different states safely and smoothly. The safety redundancy module is used to analyze the automatic driving state data and the whole vehicle driving state data, evaluate whether the current vehicle controlled state and the automatic driving classification state of the analysis module are reasonable, and force exit the automatic driving classification state control when necessary, so as to realize the direct connection between the whole vehicle control system and the automatic driving system, and bypass the energy consumption management control authority.

[0131] The automatic driving control system 402 may include a level main control module for executing different power consumption control state levels, and a power management main control module; in addition, it may also include other necessary modules required for automatic driving, which are not described in detail here. Among them, taking the example of 4 levels stored in the power consumption control state level library, then the level main control module may further include: power consumption control state level 1 main control module; power consumption control state level 2 main control module; power consumption control state level 3 main control module; power consumption control state level 4 main control module. The automatic driving control system 402 can switch between main control modules of different levels according to the decision of the automatic driving power consumption control device 401, so as to adapt the driving mode and sensors at a reasonable level to perform safe and energy-saving automatic driving control. Among them,

[0132] In power consumption control state level 1, the state monitoring module detects that the vehicle is in remote takeover mode and manual control mode, and the autonomous driving system platform and sensors are in power-on standby state. In this mode, the power consumption is the lowest.

[0133] In power control state level 2, the state detection mode detects that the vehicle is currently in autonomous driving mode. At this time, the computing unit platform of the autonomous driving system is working in a calibrated low power consumption mode (20% of full power consumption). Only the front binocular camera and the front millimeter-wave radar are working, and other sensors are in standby dormancy, realizing the autonomous driving capabilities of automatic following and autonomous cruising.

[0134] In power control state level 3, the state detection mode detects that the vehicle is currently in autonomous driving mode. At this time, the computing unit of the autonomous driving system is working at the calibrated medium power consumption (60% of the full power consumption). Only the front dual laser radar and the front monocular camera are working, and other sensors are in standby dormancy. On the basis of realizing the autonomous driving capabilities of automatic following and autonomous cruising, it can further realize the obstacle avoidance and lane change functions.

[0135] In power consumption control state level 4, the state detection mode detects that the vehicle is currently in autonomous driving mode. At this time, the autonomous driving system computing unit is operating at the calibrated full power consumption, all sensors are in working state, and all high-level autonomous driving functions are realized.

[0136] In addition, the vehicle cloud control platform 404 may be a vehicle networking platform for receiving service requests from vehicles and issuing corresponding vehicle services to the vehicles. The autonomous driving sensor system 405 may include a plurality of independent or related sensors, which are collectively regarded as an autonomous driving sensor system to provide autonomous driving services for the vehicles.

[0137] An embodiment of the present application provides a computer-readable storage medium, wherein at least one instruction is stored in the storage medium, and the at least one instruction is loaded and executed by a processor to implement the power consumption control method of the autonomous driving system as described above.

[0138] An embodiment of the present application provides an electronic device, which includes a processor and a memory, wherein the memory stores at least one instruction, and the instruction is loaded and executed by the processor to implement the power consumption control method of the autonomous driving system as described above.

[0139] An embodiment of the present application provides a computer program product, characterized in that it includes a computer program, and when the computer program is executed by a processor, it implements the power consumption control method of the autonomous driving system as described above.

[0140] An embodiment of the present application provides an autonomous driving vehicle, comprising the electronic device as described above. Specifically, the autonomous driving vehicle may be a vehicle of level L2 or above.

[0141] In the technical solution of this application, the collection, storage, use, processing, transmission, provision and disclosure of user personal information involved are in compliance with the relevant laws and regulations and do not violate public order and good morals.

[0142] Figure 5A schematic block diagram of an example electronic device 500 that can be used to implement an embodiment of the present application is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present application described and / or required herein.

[0143] like Figure 5 As shown, the electronic device 500 includes a computing unit 501, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 502 or a computer program loaded from a storage unit 508 into a random access memory (RAM) 503. In the RAM 503, various programs and data required for the operation of the electronic device 500 can also be stored. The computing unit 501, the ROM 502, and the RAM 503 are connected to each other via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0144] Multiple components in the electronic device 500 are connected to the I / O interface 505, including: an input unit 506, such as a keyboard, a mouse, etc.; an output unit 507, such as various types of displays, speakers, etc.; a storage unit 508, such as a disk, an optical disk, etc.; and a communication unit 509, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 509 allows the electronic device 500 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0145] The computing unit 501 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 501 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 501 performs the various methods and processes described above, such as the power consumption control method of the autonomous driving system. For example, in some embodiments, the power consumption control method of the autonomous driving system may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 508. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 500 via the ROM 502 and / or the communication unit 509. When the computer program is loaded into the RAM 503 and executed by the computing unit 501, one or more steps of the power consumption control method of the autonomous driving system described above may be performed. Alternatively, in other embodiments, the computing unit 501 may be configured to execute the power consumption control method of the autonomous driving system in any other appropriate manner (e.g., by means of firmware).

[0146] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0147] The program code for implementing the method of the present application can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, implements the functions / operations specified in the flow chart and / or block diagram. The program code can be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0148] In the context of the present application, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, 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 foregoing.

[0149] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0150] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0151] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship with each other. The server may be a cloud server, a server of a distributed system, or a server combined with a blockchain.

[0152] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps disclosed in this application can be performed in parallel, sequentially or in different orders, as long as the desired results of the technical solution disclosed in this application can be achieved, and this document does not limit this.

[0153] The above specific implementations do not constitute a limitation on the protection scope of this application. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of this application should be included in the protection scope of this application.

Claims

1. A method for controlling power consumption of an automatic driving system, characterized in that: include: Obtain road information and vehicle information of the target vehicle in the current driving environment; Extracting road scene features from the road information and extracting vehicle posture control features from the vehicle information, and based on the road scene features and the vehicle posture control features, using the trained share prediction model, predicting the sensor demand share of the target vehicle in the current road scene; According to the predicted demand ratio, searching for a matching power consumption control state level from a preset power consumption control state classification library; wherein the preset power consumption control state classification library stores a plurality of power consumption control state levels, each power consumption control state level being configured with a driving mode, a demand ratio, and an operable sensor identifier; Based on the driving mode and the operable sensor identifiers in the found power consumption control state level, a target driving mode and a target sensor are determined respectively, the target vehicle is controlled to operate the target sensor, and automatic driving control is performed in the target driving mode.

2. The method according to claim 1, characterized in that Based on the road scene features and the vehicle posture control features, the trained share prediction model is used to predict the sensor demand share of the target vehicle in the current road scene, specifically including: The road scene features and the vehicle posture control features are respectively input into the trained proportion prediction model to predict the proportion of sensor demand of the target vehicle in the current road scene; wherein the trained proportion prediction model is a multimodal prediction model.

3. The method according to claim 2, characterized in that The proportion prediction model is trained in the following way: Obtaining historical data of autonomous driving of sample vehicles of the same model as the target vehicle when running at full power consumption in various road scenarios; A sample data set for training is selected from the autonomous driving historical data; wherein the sample data set includes a plurality of sample data and a sample label corresponding to each sample data, and each sample data includes at least: a road scene feature and a vehicle posture control feature; and the sample label is an operating ratio of a plurality of sensors on the sample vehicle; Inputting a plurality of sample data consisting of road scene features and vehicle posture control features and sample labels in the sample data set into a preset neural network model for training; Based on the optimal control strategy, the model parameters are repeatedly adjusted to obtain a converged share prediction model.

4. The method according to any one of claims 1 to 3, characterized in that: The method further comprises: Periodically detect the road information of the target vehicle in the current driving environment; If a change in road information is detected, the latest road information and vehicle information of the target vehicle in the current driving environment are re-acquired; and based on the latest road information and vehicle information, the road scene feature and vehicle posture control feature are respectively extracted, and the trained proportion prediction model is used to predict the latest demand proportion of the target vehicle for the sensor in the current road scene; According to the predicted latest demand ratio, searching for the matching latest power consumption control state level from a preset power consumption control state classification library; The target vehicle is controlled to switch from a current power consumption control state level to a latest power consumption control state level.

5. The method according to any one of claims 1 to 3, characterized in that: The method further comprises: Periodically receive vehicle control data and autonomous driving control data and analyze them; Evaluate whether the power consumption control state level of the target vehicle in the current driving environment matches according to the analysis result; If they do not match, the power consumption control state level is exited, triggering entry into the direct control mode between the vehicle control system and the autonomous driving system.

6. A power consumption control device for an automatic driving system, characterized in that: include: An acquisition module is used to acquire road information and vehicle information of a target vehicle in a current driving environment; A prediction module, used to extract road scene features from the road information, and extract vehicle posture control features from the vehicle information, and based on the road scene features and the vehicle posture control features, use the trained proportion prediction model to predict the proportion of sensor demand of the target vehicle in the current road scene; A search module, configured to search for a matching power consumption control state level from a preset power consumption control state classification library according to the predicted demand proportion; wherein the preset power consumption control state classification library stores a plurality of power consumption control state levels, each power consumption control state level being configured with a driving mode, a demand proportion, and an operable sensor identifier; The control module is used to determine the target driving mode and the target sensor respectively based on the driving mode and the operable sensor identifier in the power consumption control state level found, control the target vehicle to operate the target sensor, and perform automatic driving control in the target driving mode.

7. An electronic device, characterized in that: include: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 5.

8. A non-transitory computer-readable storage medium storing computer instructions, characterized in that: The computer instructions are used to cause the computer to execute the method according to any one of claims 1-5.

9. A computer program product, characterized in that The invention comprises a computer program which, when executed by a processor, implements the method according to any one of claims 1 to 5.

10. An autonomous driving vehicle, characterized in that: Comprising the electronic device as claimed in claim 7.

Images