Intelligent monitoring system and method for electronic control unit of pure electric heavy truck

Abstract

The invention discloses an intelligent monitoring system and method for an electronic control unit of a pure electric heavy truck, and relates to the technical field of intelligent monitoring, and the method comprises the steps: monitoring and obtaining a temperature sequence of a power electronic control unit of the pure electric heavy truck, and collecting the heavy truck feature information of the pure electric heavy truck; a front road image is obtained, road condition recognition is carried out, and road condition information and road condition precision are obtained; according to the road condition precision and the heavy truck feature information, configuring a temperature influence prediction scale for temperature influence prediction, and according to the temperature influence prediction scale, the heavy truck feature information, the road condition information and the temperature sequence, performing temperature influence prediction to obtain a predicted temperature; according to the predicted temperature, state parameters of the power electronic control unit are obtained in a classified mode and serve as monitoring results. According to the invention, the technical problems of inaccurate monitoring of the running state of the power electronic control unit and lack of fault early warning capability in the prior art are solved, and the technical effects of real-time monitoring of the running state of the equipment and fault early warning are realized.

Description

Technical Field

[0001] The present invention relates to the field of intelligent monitoring technology, and in particular to an intelligent monitoring system and method for an electronic control unit of a pure electric heavy truck. Background Art

[0002] With the development of new energy technologies and the rapid upgrading of the transportation industry, pure electric heavy-duty trucks have been widely used in logistics and transportation due to their high efficiency and environmental protection. However, with the diversification of transportation tasks and the complexity of operating conditions, the monitoring and management of the operating status of the power electronic control unit (ECU) of pure electric heavy-duty trucks, as a core component, are also facing increasing challenges. Traditional power electronic control unit monitoring methods often have problems such as inaccurate monitoring of the operating status of the power electronic control unit and lack of fault warning capabilities, which makes it difficult to meet the modern transportation industry's demand for efficient and safe operation of pure electric heavy-duty trucks. Summary of the invention

[0003] The present application provides a system and method for intelligent monitoring of the electronic control unit of a pure electric heavy truck, which is used to solve the technical problems of the prior art in inaccurate monitoring of the operating status of the power electronic control unit and lack of fault warning capability.

[0004] In view of the above problems, the present application provides an intelligent monitoring system and method for an electronic control unit of a pure electric heavy truck.

[0005] In a first aspect of the present application, a pure electric heavy-duty truck electronic control unit intelligent monitoring system is provided, the system comprising:

[0006] A data monitoring and acquisition module, the data monitoring and acquisition module is used to monitor and obtain the temperature sequence of the power electronic control unit of the pure electric heavy-duty truck, and collect the heavy-duty truck characteristic information of the pure electric heavy-duty truck; a road condition recognition module, the road condition recognition module is used to obtain the image of the road ahead, perform road condition recognition, and obtain road condition information and road condition accuracy; a temperature impact prediction module, the temperature impact prediction module is used to configure a temperature impact prediction scale for temperature impact prediction according to the road condition accuracy and heavy-duty truck characteristic information, and perform temperature impact prediction according to the temperature impact prediction scale, the heavy-duty truck characteristic information, the road condition information and the temperature sequence to obtain a predicted temperature; a monitoring result acquisition module, the monitoring result acquisition module is used to classify and obtain the state parameters of the power electronic control unit according to the predicted temperature as the monitoring result.

[0007] A second aspect of the present application provides a method for intelligent monitoring of an electronic control unit of a pure electric heavy truck, the method comprising:

[0008] Monitor and obtain the temperature sequence of the power electronic control unit of the pure electric heavy-duty truck, and collect the heavy-duty truck characteristic information of the pure electric heavy-duty truck; obtain the image of the road ahead, perform road condition recognition, and obtain road condition information and road condition accuracy; configure a temperature impact prediction scale for temperature impact prediction based on the road condition accuracy and heavy-duty truck characteristic information, perform temperature impact prediction according to the temperature impact prediction scale, based on the heavy-duty truck characteristic information, road condition information and temperature sequence, and obtain a predicted temperature; and classify and obtain the state parameters of the power electronic control unit based on the predicted temperature as a monitoring result.

[0009] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0010] The present application monitors and obtains the temperature sequence of the power electronic control unit of a pure electric heavy truck, and collects the heavy truck characteristic information of the pure electric heavy truck; obtains the image of the road ahead, performs road condition identification, and obtains road condition information and road condition accuracy; configures the temperature impact prediction scale for temperature impact prediction according to the road condition accuracy and heavy truck characteristic information, and performs temperature impact prediction according to the temperature impact prediction scale, the heavy truck characteristic information, road condition information and temperature sequence, and obtains the predicted temperature; and classifies and obtains the state parameters of the power electronic control unit according to the predicted temperature as the monitoring result. The present invention solves the technical problems of the prior art in that the operating status of the power electronic control unit is not accurately monitored and the fault warning capability is lacking. By monitoring the temperature sequence and heavy truck characteristic information, identifying the road condition information and accuracy, and dynamically configuring the temperature impact prediction scale, the temperature change of the power electronic control unit is accurately predicted and the operating status is output by classification, combining the temperature impact prediction and classification model, so as to achieve the technical effect of real-time monitoring of the equipment operating status and fault warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0012] Figure 1 A schematic diagram of the structure of an intelligent monitoring system for an electronic control unit of a pure electric heavy truck provided in an embodiment of the present application;

[0013] Figure 2 A schematic flow chart of an intelligent monitoring method for an electronic control unit of a pure electric heavy truck provided in an embodiment of the present application.

[0014] Explanation of the reference numerals: data monitoring and collecting module 11 , road condition identification module 12 , temperature impact prediction module 13 , monitoring result acquisition module 14 . DETAILED DESCRIPTION

[0015] The present application provides an intelligent monitoring system and method for the electronic control unit of a pure electric heavy-duty truck, aiming to solve the technical problems of the prior art in inaccurate monitoring of the operating status of the power electronic control unit and lack of fault warning capability. By monitoring the temperature sequence and characteristic information of the heavy-duty truck, identifying the road condition information and accuracy, and dynamically configuring the temperature impact prediction scale, the temperature change of the power electronic control unit is accurately predicted and the operating status is output by classification, thereby achieving the technical effect of real-time monitoring of the equipment operating status and fault warning.

[0016] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0017] It should be noted that any variations of the terms "include" and "have" are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules that are not explicitly listed or inherent to these processes, methods, products or devices.

[0018] Embodiment 1, as Figure 1 As shown, the embodiment of the present application provides an intelligent monitoring system for an electronic control unit of a pure electric heavy truck, the system comprising:

[0019] The data monitoring and acquisition module 11 is used to monitor and acquire the temperature sequence of the power electronic control unit of the pure electric heavy-duty truck, and to acquire the heavy-duty truck characteristic information of the pure electric heavy-duty truck.

[0020] In the embodiment of the present application, the data monitoring and acquisition module first monitors the temperature data of multiple time points in the past through the temperature sensor on the power electronic control unit to generate a complete temperature sequence. These temperature data reflect the thermal performance change trend of the power electronic control unit under different working conditions.

[0021] Then, the current load and continuous mileage of the heavy truck are collected in real time. The load is obtained through the on-board weighing sensor to represent the current load of the heavy truck, and the continuous mileage is recorded by the odometer to reflect the continuous mileage of the vehicle's current operation. In addition, the cumulative mileage of the vehicle is extracted. This data is generated from the vehicle's operation history and is used to measure the long-term use of the power system. Finally, the load, continuous mileage and cumulative mileage are integrated into the characteristic information of the heavy truck.

[0022] Furthermore, in the system provided in the embodiment of the application, the data monitoring and collection module 11 is also used for:

[0023] Monitor and obtain the temperature of the power electronic control unit of the pure electric heavy-duty truck at multiple times in the past to obtain a temperature sequence; collect the current load and continuous mileage of the pure electric heavy-duty truck; collect the cumulative mileage of the pure electric heavy-duty truck; integrate the load, continuous mileage and cumulative mileage to obtain heavy-duty truck characteristic information.

[0024] In the embodiment of the present application, a high-precision temperature sensor (such as a thermocouple or RTD) installed on the surface of the power electronic control unit is first used to collect temperature data in real time. The temperature data collection time interval can be set to 1 minute to dynamically capture the temperature change trend of the power electronic control unit, forming a temperature sequence containing 5 time points to describe its thermal characteristics changes in the past five minutes. Among them, the temperature data collection time interval can be set according to needs.

[0025] Then the current load and continuous driving mileage of the pure electric heavy truck are collected. The load is measured in real time by the on-board weighing sensor installed on the vehicle chassis, reflecting the current load situation, which is a direct factor affecting the vehicle's power consumption and system temperature; the continuous driving mileage is accumulated in real time by the on-board odometer module (based on Hall sensor or photoelectric encoder), which is continuously recorded from the start to the stop of the vehicle, reflecting the current distance traveled by the vehicle.

[0026] In addition, the accumulated mileage is extracted from the operation data recording unit (EDR) of the vehicle controller. By synchronously storing the mileage data of each vehicle trip in the EDR, the total distance traveled by the vehicle since its operation is counted as an important indicator of the long-term use status of the power system.

[0027] Finally, the load capacity, continuous mileage and cumulative mileage are integrated to generate the heavy truck characteristic information.

[0028] The road condition recognition module 12 is used to obtain a front road image, perform road condition recognition, and obtain road condition information and road condition accuracy.

[0029] In the embodiment of the present application, the road condition recognition module first uses the image acquisition device of the pure electric heavy truck to obtain image data of the road ahead, and analyzes the image through a pre-trained road condition identifier to extract road condition information.

[0030] Then, the brightness information of the image is extracted, compared with the preset brightness, and the ratio is calculated as the first road condition accuracy. At the same time, the current visibility data is collected, compared with the preset visibility, and the ratio is calculated as the second road condition accuracy. Finally, the average of the first road condition accuracy and the second road condition accuracy is output as the road condition accuracy.

[0031] Furthermore, in the system provided in the embodiment of the application, the road condition recognition module 12 is also used for:

[0032] A road image of the road ahead is collected by an image acquisition device of a pure electric heavy truck; road condition recognition is performed on the road image to obtain road condition information; brightness extraction is performed on the road image to obtain brightness information, and a ratio of the brightness information to a preset brightness is calculated as a first road condition accuracy; current visibility is obtained, and a ratio to the preset visibility is calculated as a second road condition accuracy; an average of the first road condition accuracy and the second road condition accuracy is calculated to obtain road condition accuracy.

[0033] In an embodiment of the present application, firstly, an image acquisition device (such as a high-definition vehicle-mounted camera or a laser radar) installed at the front end of the pure electric heavy truck is used to acquire a road image of the road ahead in real time.

[0034] Next, the collected road images are used for road condition recognition. Specifically, a set of sample road images is first collected and features such as road slope in the images are annotated to generate a set of sample intersection information. Then, the network architecture of the road condition recognition device is constructed based on the convolutional neural network, and the model is supervised, trained, verified, and tested using the above data, ultimately achieving road condition recognition of actual road images and extracting road condition information.

[0035] After the road condition information is extracted, the brightness information of the road image is extracted. Brightness information is an important parameter for measuring road visibility. The road image is converted into a grayscale image, the grayscale value of each pixel is counted, the average brightness of the image is calculated, and the ratio is calculated with the preset brightness (i.e. the maximum brightness under natural light during driving, usually set to a grayscale value of 250) to obtain the first road condition accuracy.

[0036] At the same time, the vehicle's current geographic location is obtained by combining the on-board GPS positioning, and the weather data of the location is obtained in real time by connecting to the meteorological service platform. The current visibility is extracted and compared with the preset visibility (the maximum visibility under sunny conditions is 5 kilometers) to obtain the second road condition accuracy, which is used to measure the visual conditions of the driving environment.

[0037] Finally, the first road condition accuracy and the second road condition accuracy are averaged to obtain the final road condition accuracy.

[0038] Furthermore, in the system provided in the embodiment of the application, the road condition recognition module 12 is also used for:

[0039] A set of sample road images is collected, and the road slope in each sample road image is marked to obtain a set of sample intersection information; a network architecture of a road condition identifier is constructed based on a convolutional neural network; the sample road image set and the sample intersection information set are used as supervised training data to perform supervised training, verification and testing on the road condition identifier to obtain road condition information.

[0040] In the embodiment of the present application, a large number of sample road images are first acquired through historical database collection to form a sample road image set. The images cover different time, weather and light conditions to ensure data diversity, including common scenes such as straight roads, uphill and downhill.

[0041] Then, professional annotation tools (such as CVAT) are used to annotate each sample road image in detail, marking the slope information of the road in the image, such as recording the slope as 0° (flat road), 5° (small slope), 10° (steep slope), etc. These annotation data form a sample road condition information set, which corresponds one-to-one to the road images in the sample road image set.

[0042] Next, we build the network architecture of the road condition identifier based on the convolutional neural network (CNN). The network consists of multiple key layers, including the convolution layer for extracting low-level features (such as edges and textures) and high-level features (such as road morphology and slope changes) in the image; the pooling layer reduces the data dimension through the maximum pooling operation while retaining key features; the fully connected layer maps the extracted features to specific road slope values; and the output layer outputs the prediction results in the form of regression, that is, the specific slope angle value corresponding to the image.

[0043] The sample road image set and the sample road condition information set are then divided into a training set, a validation set, and a test set for model training, validation, and testing, respectively. In the training phase, a mapping relationship is established between the input image and the annotated slope value, and the mean square error (MSE) is used as the loss function to calculate the difference between the model prediction value and the true value. The Adam optimizer is used to dynamically adjust the network weights to reduce the loss. In the validation phase, an independent validation set is used to evaluate the generalization ability of the model, and the model performance is optimized by adjusting hyperparameters (such as learning rate and network depth). In the testing phase, an independent test set is used to calculate the mean square error (MSE) and determination coefficient (R²) to verify the prediction accuracy of the model on actual data.

[0044] After training is completed, the actual collected road images are input into the trained road condition identifier. The road condition identifier extracts image features through the convolution layer, maps these features to the learned slope information through the fully connected layer, and finally outputs the specific slope angle of the road (such as 0°, 5°, 10°, etc.) in the form of regression in the output layer. This output is the road condition information.

[0045] The temperature impact prediction module 13 is used to configure a temperature impact prediction scale for temperature impact prediction according to the road condition accuracy and heavy truck characteristic information, and to perform temperature impact prediction according to the temperature impact prediction scale, the heavy truck characteristic information, the road condition information and the temperature sequence to obtain a predicted temperature.

[0046] In the embodiment of the present application, the temperature impact prediction module dynamically configures the temperature impact prediction scale by integrating the road condition accuracy and heavy truck characteristic information, and combines the historical temperature sequence, heavy truck characteristic information and road condition information to predict the temperature change of the power electronic control unit. Specifically, firstly, the maximum load, maximum continuous mileage and maximum cumulative mileage are compared with the current heavy truck characteristic information, and the ratio is calculated and the average is taken to obtain the driving recognition accuracy. Next, combined with the road condition accuracy, the temperature impact prediction coefficient is obtained by subtracting the difference between the road condition accuracy and the driving recognition accuracy from 1 and averaging it, as the temperature impact prediction scale.

[0047] In the prediction stage, we first collect and construct a set of sample temperature sequences, a set of sample heavy truck feature information, a set of sample road condition information, and a set of sample predicted temperature. These data are divided into K training data. Through supervised training, K temperature impact prediction paths are generated and combined to form a temperature impact prediction channel. Next, according to the temperature impact prediction scale, the number of prediction paths to be selected is dynamically determined, that is, by multiplying the temperature impact prediction scale by K and then rounding up to get the number of paths to be extracted. Then, the corresponding number of paths are randomly selected from the temperature impact prediction channel, and the current heavy truck feature information, road condition information, and temperature sequence are input into the selected path for prediction. Finally, the output temperature values ​​of multiple prediction paths are combined and the average is calculated to obtain the final predicted temperature.

[0048] Furthermore, in the system provided in the embodiment of the application, the temperature impact prediction module 13 is also used for:

[0049] The maximum load capacity, the maximum continuous driving mileage and the maximum cumulative driving mileage are obtained; the ratios of the load capacity, the continuous driving mileage and the cumulative driving mileage in the heavy truck characteristic information to the maximum load capacity, the maximum continuous driving mileage and the maximum cumulative driving mileage are calculated respectively, and the average is calculated to obtain the driving recognition accuracy; according to the driving recognition accuracy and the road condition accuracy, the temperature impact prediction coefficient is calculated as the temperature impact prediction scale.

[0050] In an embodiment of the present application, the maximum load weight, maximum continuous driving mileage and maximum cumulative driving mileage are first obtained from the vehicle's design database or electronic control unit (ECU). These parameters respectively represent the upper limit of the vehicle's carrying capacity under design performance, the theoretical maximum endurance on a single charge and the maximum driving distance over the vehicle's entire life cycle.

[0051] The current operating status data is then collected in real time through on-board sensors and the vehicle management system (VMS), including the current load (measured in real time by the pressure sensor in the suspension system or the on-board weighing module), the current continuous mileage (the odometer value recorded by the photoelectric encoder or Hall sensor), and the current cumulative mileage (the historical cumulative value is queried by the operation data recorder EDR). By calculating the ratio of the above real-time data with the corresponding maximum parameter value, the load ratio, continuous mileage ratio, and cumulative mileage ratio are obtained. The driving recognition accuracy is obtained by calculating the average of the load ratio, continuous mileage ratio, and cumulative mileage ratio.

[0052] Next, the temperature impact prediction coefficient is calculated based on the driving recognition accuracy and road condition accuracy. Specifically, the difference between 1 and the road condition accuracy is first calculated, and then the average of the difference and the driving recognition accuracy obtained above is calculated to obtain the temperature impact prediction coefficient. Finally, the temperature impact prediction coefficient is used as the temperature impact prediction scale.

[0053] Furthermore, in the system provided in the embodiment of the application, the temperature impact prediction module 13 is also used for:

[0054] According to the historical temperature monitoring data of the power electronic control unit of the pure electric heavy-duty truck, a sample temperature sequence set, a sample heavy-duty truck characteristic information set, a sample road condition information set, and a sample predicted temperature set are collected; the sample temperature sequence set, the sample heavy-duty truck characteristic information set, the sample road condition information set, and the sample predicted temperature set are divided into K portions of temperature impact prediction supervision training data, where K is a positive integer; the K portions of temperature impact prediction supervision training data are used to train to obtain K temperature impact prediction paths, which are combined to obtain a temperature impact prediction channel; according to the temperature impact prediction scale and the K temperature impact prediction paths, at least one temperature impact prediction path is calculated and selected, the heavy-duty truck characteristic information, road condition information, and temperature sequence are respectively input, at least one predicted temperature is predicted, and the mean is calculated to obtain the final predicted temperature.

[0055] In the embodiment of the present application, firstly, based on the historical temperature monitoring data of the power electronic control unit, a sample temperature sequence set, a sample heavy truck feature information set, a sample road condition information set and a sample predicted temperature set are collected and constructed. Among them, the sample temperature sequence set contains temperature records at multiple time points, reflecting the temperature change trend of the power electronic control unit within a specific time window (such as the past 5 minutes); the sample heavy truck feature information set records the vehicle operating status parameters, including load capacity, continuous mileage and cumulative mileage, etc.; the sample road condition information set describes the road environment characteristics, such as road slope; the sample predicted temperature set uses the actual temperature 1 minute after the end of the temperature sequence as the target value, which is used as the supervision label for subsequent model training.

[0056] After data collection is completed, the four sample sets are divided into K pieces of supervised training data for temperature impact prediction, where K is a positive integer. Each piece of data contains a temperature sequence, heavy truck feature information, road condition information, and the corresponding predicted temperature, ensuring the independence and diversity of the data and improving the generalization ability of model training. Then, based on these supervised training data, deep learning techniques (such as LSTM long short-term memory network) are used to train K temperature impact prediction paths. Specifically, a path is trained independently for each piece of data. The LSTM model captures the time dependency of the temperature sequence through its memory unit, combines the heavy truck features and road condition information, and learns the complex mapping relationship between the input features and the predicted temperature, thereby generating a unique prediction logic. Finally, K independent paths are combined to form a temperature impact prediction channel.

[0057] In the prediction stage, the number of paths to be selected is determined based on the calculated temperature impact prediction scale. The number of paths is obtained by multiplying the temperature impact prediction scale by K and rounding up to ensure the dynamic and adaptable nature of path selection. For example, if the temperature impact prediction scale is 0.7 and K is 10, 7 paths need to be selected for prediction. The corresponding number of paths are randomly selected from the prediction channel, and each path independently outputs a predicted temperature value.

[0058] After the path selection is completed, the heavy truck feature information, road condition information and real-time temperature sequence in the current running state are input into the selected prediction path, and the predicted temperature value of each path is calculated independently. Then, the predicted temperature values ​​output by all paths are averaged to obtain the final predicted temperature.

[0059] The monitoring result acquisition module 14 is used to classify and obtain the state parameters of the power electronic control unit according to the predicted temperature as the monitoring result.

[0060] In an embodiment of the present application, the monitoring result acquisition module collects a sample temperature set of the power electronic control unit, and labels the corresponding state level for different temperatures, constructs a mapping relationship between temperature and state, and generates a state monitoring classifier. The state level is labeled based on the negative correlation between the temperature and the operating state of the equipment. For example, the higher the temperature, the lower the state level. The predicted temperature obtained above is input into the state monitoring classifier, and the real-time state parameters of the power electronic control unit are obtained through rapid classification through the mapping relationship as the monitoring result output.

[0061] Furthermore, in the system provided in the embodiment of the application, the monitoring result acquisition module 14 is also used for:

[0062] A sample temperature set of the power electronic control unit is collected, and different state levels are marked for different sample temperatures to obtain a sample state parameter set, wherein the state level is negatively correlated with the sample temperature; a mapping relationship between the sample temperature set and the sample state parameter set is constructed to obtain a state monitoring classifier; the predicted temperature is input into the state monitoring classifier, and the state parameters of the power electronic control unit are obtained by mapping and classification as monitoring results.

[0063] In an embodiment of the present application, a sample temperature set of the power electronic control unit is first collected. The sample temperature set is obtained through historical operating data or real-time collection of the power electronic control unit, covering the temperature range of the power electronic control unit under different operating conditions, such as low load, normal operation, high load and critical state. Specifically, the sample temperature data can be recorded in real time by the temperature sensor built into the power electronic control unit and stored in an electronic data recorder (EDR) or a cloud database. The sample temperature set includes key temperature points, such as 90°C, 70°C, 50°C, and 30°C. These temperature values ​​fully reflect the thermal characteristics of the equipment under high load, normal operation, and low load.

[0064] After completing the collection of the sample temperature set, different state levels are marked for different sample temperatures according to the design specifications and operating experience of the power electronic control unit to generate a sample state parameter set. The state level is expressed by a numerical value (such as level 1, level 2, level 3, and level 4), which is used to indicate the operating state of the power electronic control unit. The size of the state level is negatively correlated with the size of the sample temperature, that is, the higher the sample temperature, the smaller the state level. Specifically, level 1 corresponds to the critical state, the sample temperature is the highest (such as 90°C and above), indicating that the power electronic control unit is in a dangerous state and may need to stop running immediately. Level 2 corresponds to the overheating state, the sample temperature is high (such as between 70°C and 90°C), and attention should be paid to load reduction or heat dissipation. Level 3: corresponds to the normal operating state, the sample temperature is moderate (such as between 50°C and 70°C), and it is in a good working state. Level 4 corresponds to the ideal state, the sample temperature is the lowest (such as below 50°C), indicating that the equipment is operating under optimal conditions. Through this marking rule, the sample temperature set and the sample state parameter set form a corresponding relationship.

[0065] After the state labeling is completed, a one-to-one mapping relationship is established between the sample temperature set and the sample state parameter set to construct a state monitoring classifier. The classifier quickly converts the temperature value into the corresponding state level through the preset mapping rules without the need for complex calculation processes. In actual operation, the state monitoring classifier is essentially a table lookup tool that can map the predicted temperature to the corresponding state level through a simple search operation, thereby intuitively reflecting the current operating state of the equipment.

[0066] When the predicted temperature of the power electronic control unit is provided by the temperature impact prediction module, the predicted temperature is input into the condition monitoring classifier. The condition monitoring classifier quickly outputs the corresponding state parameters by looking up the mapping rules. For example, when the predicted temperature is 75°C, the classifier will map it to "level 2", indicating that the device is in an overheated state; if the predicted temperature is 95°C, the classifier will map it to "level 1", indicating that the device has entered a critical state.

[0067] Finally, the state parameters output by the state monitoring classifier are used as monitoring results to provide a reference for users' driving.

[0068] In the embodiments of the present application, in summary, the embodiments of the present application have at least the following technical effects:

[0069] The present application monitors and obtains the temperature sequence of the power electronic control unit of a pure electric heavy truck, and collects the heavy truck characteristic information of the pure electric heavy truck; obtains the image of the road ahead, performs road condition identification, and obtains road condition information and road condition accuracy; configures the temperature impact prediction scale for temperature impact prediction according to the road condition accuracy and heavy truck characteristic information, and performs temperature impact prediction according to the temperature impact prediction scale, the heavy truck characteristic information, road condition information and temperature sequence, and obtains the predicted temperature; and classifies and obtains the state parameters of the power electronic control unit according to the predicted temperature as the monitoring result. The present invention solves the technical problems of the prior art in that the operating status of the power electronic control unit is not accurately monitored and the fault warning capability is lacking. By monitoring the temperature sequence and heavy truck characteristic information, identifying the road condition information and accuracy, and dynamically configuring the temperature impact prediction scale, the temperature change of the power electronic control unit is accurately predicted and the operating status is output by classification, combining the temperature impact prediction and classification model, so as to achieve the technical effect of real-time monitoring of the equipment operating status and fault warning.

[0070] Embodiment 2 is based on the same inventive concept as the above embodiment of a pure electric heavy truck electronic control unit intelligent monitoring system. Figure 2 As shown, the embodiment of the present application provides a method for intelligent monitoring of an electronic control unit of a pure electric heavy truck, the method comprising:

[0071] Monitor and obtain the temperature sequence of the power electronic control unit of the pure electric heavy-duty truck, and collect the heavy-duty truck characteristic information of the pure electric heavy-duty truck; obtain the image of the road ahead, perform road condition recognition, and obtain road condition information and road condition accuracy; configure a temperature impact prediction scale for temperature impact prediction based on the road condition accuracy and heavy-duty truck characteristic information, perform temperature impact prediction according to the temperature impact prediction scale, based on the heavy-duty truck characteristic information, road condition information and temperature sequence, and obtain a predicted temperature; and classify and obtain the state parameters of the power electronic control unit based on the predicted temperature as a monitoring result.

[0072] Further, the temperature sequence of the power electronic control unit of the pure electric heavy-duty truck is monitored and acquired, and the heavy-duty truck characteristic information of the pure electric heavy-duty truck is collected. The method further includes:

[0073] Monitor and obtain the temperature of the power electronic control unit of the pure electric heavy-duty truck at multiple times in the past to obtain a temperature sequence; collect the current load and continuous mileage of the pure electric heavy-duty truck; collect the cumulative mileage of the pure electric heavy-duty truck; integrate the load, continuous mileage and cumulative mileage to obtain heavy-duty truck characteristic information.

[0074] Furthermore, the method further includes acquiring a road image ahead, performing road condition recognition, and obtaining road condition information and road condition accuracy:

[0075] A road image of the road ahead is collected by an image acquisition device of a pure electric heavy truck; road condition recognition is performed on the road image to obtain road condition information; brightness extraction is performed on the road image to obtain brightness information, and a ratio of the brightness information to a preset brightness is calculated as a first road condition accuracy; current visibility is obtained, and a ratio to the preset visibility is calculated as a second road condition accuracy; an average of the first road condition accuracy and the second road condition accuracy is calculated to obtain road condition accuracy.

[0076] Furthermore, the road condition recognition is performed on the road image to obtain road condition information, and the method further includes:

[0077] A set of sample road images is collected, and the road slope in each sample road image is marked to obtain a set of sample intersection information; a network architecture of a road condition identifier is constructed based on a convolutional neural network; the sample road image set and the sample intersection information set are used as supervised training data to perform supervised training, verification and testing on the road condition identifier to obtain road condition information.

[0078] Further, according to the road condition accuracy and the heavy truck feature information, a temperature impact prediction scale for performing temperature impact prediction is configured, and the method further includes:

[0079] The maximum load capacity, the maximum continuous driving mileage and the maximum cumulative driving mileage are obtained; the ratios of the load capacity, the continuous driving mileage and the cumulative driving mileage in the heavy truck characteristic information to the maximum load capacity, the maximum continuous driving mileage and the maximum cumulative driving mileage are calculated respectively, and the average is calculated to obtain the driving recognition accuracy; according to the driving recognition accuracy and the road condition accuracy, the temperature impact prediction coefficient is calculated as the temperature impact prediction scale.

[0080] Further, according to the temperature impact prediction scale, based on the heavy truck feature information, road condition information and temperature sequence, temperature impact prediction is performed to obtain a predicted temperature, and the method further includes:

[0081] According to the historical temperature monitoring data of the power electronic control unit of the pure electric heavy-duty truck, a sample temperature sequence set, a sample heavy-duty truck characteristic information set, a sample road condition information set, and a sample predicted temperature set are collected; the sample temperature sequence set, the sample heavy-duty truck characteristic information set, the sample road condition information set, and the sample predicted temperature set are divided into K portions of temperature impact prediction supervision training data, where K is a positive integer; the K portions of temperature impact prediction supervision training data are used to train to obtain K temperature impact prediction paths, which are combined to obtain a temperature impact prediction channel; according to the temperature impact prediction scale and the K temperature impact prediction paths, at least one temperature impact prediction path is calculated and selected, the heavy-duty truck characteristic information, road condition information, and temperature sequence are respectively input, at least one predicted temperature is predicted, and the mean is calculated to obtain the final predicted temperature.

[0082] Further, according to the predicted temperature, the state parameters of the power electronic control unit are obtained by classification as monitoring results, and the method further includes:

[0083] A sample temperature set of the power electronic control unit is collected, and different state levels are marked for different sample temperatures to obtain a sample state parameter set, wherein the state level is negatively correlated with the sample temperature; a mapping relationship between the sample temperature set and the sample state parameter set is constructed to obtain a state monitoring classifier; the predicted temperature is input into the state monitoring classifier, and the state parameters of the power electronic control unit are obtained by mapping and classification as monitoring results.

[0084] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. The processes depicted in the accompanying drawings do not necessarily require the specific order and continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0085] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

[0086] This specification and drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. An intelligent monitoring system for an electronic control unit of a pure electric heavy truck, characterized in that: The system comprises: A data monitoring and acquisition module, the data monitoring and acquisition module is used to monitor and acquire the temperature sequence of the power electronic control unit of the pure electric heavy-duty truck, and to collect the heavy-duty truck characteristic information of the pure electric heavy-duty truck; A road condition recognition module, which is used to obtain a road image ahead, perform road condition recognition, and obtain road condition information and road condition accuracy; A temperature impact prediction module, the temperature impact prediction module is used to configure a temperature impact prediction scale for temperature impact prediction according to the road condition accuracy and the heavy truck characteristic information, and perform temperature impact prediction according to the temperature impact prediction scale, the heavy truck characteristic information, the road condition information and the temperature sequence to obtain a predicted temperature; A monitoring result acquisition module is used to classify and obtain the state parameters of the power electronic control unit as monitoring results according to the predicted temperature.

2. The intelligent monitoring system for the electronic control unit of a pure electric heavy truck according to claim 1 is characterized in that: Data monitoring and acquisition module, used for: Monitor and obtain the temperature of the power electronic control unit of a pure electric heavy truck at multiple times in the past to obtain a temperature sequence; Collecting the current load capacity and continuous driving mileage of the pure electric heavy truck; Collecting the accumulated mileage of the pure electric heavy truck; The load capacity, continuous mileage and cumulative mileage are integrated to obtain heavy truck characteristic information.

3. The intelligent monitoring system for the electronic control unit of a pure electric heavy truck according to claim 1 is characterized in that: Road condition recognition module, used for: The image acquisition device of the pure electric heavy truck is used to collect the road image of the road ahead; Performing road condition recognition on the road image to obtain road condition information; Performing brightness extraction on the road image to obtain brightness information, and calculating a ratio between the brightness information and a preset brightness as a first road condition accuracy; Obtaining the current visibility and calculating the ratio to the preset visibility as the second road condition accuracy; The average of the first road condition accuracy and the second road condition accuracy is calculated to obtain the road condition accuracy.

4. The intelligent monitoring system for the electronic control unit of a pure electric heavy truck according to claim 3 is characterized in that: Road condition recognition module, used for: Collecting a set of sample road images, marking the road slope in each sample road image, and obtaining a set of sample intersection information; Based on convolutional neural network, the network architecture of road condition recognition is constructed; The sample road image set and the sample intersection information set are used as supervised training data to perform supervised training, verification and testing on the road condition identifier to obtain road condition information.

5. The intelligent monitoring system for the electronic control unit of a pure electric heavy truck according to claim 2 is characterized in that: Temperature impact prediction module for: Obtain the maximum load capacity, maximum continuous mileage and maximum cumulative mileage; Respectively calculating the ratio of the load, continuous driving mileage and cumulative driving mileage in the heavy truck characteristic information to the maximum load, maximum continuous driving mileage and maximum cumulative driving mileage, and calculating the average to obtain the driving recognition accuracy; According to the driving recognition accuracy and the road condition accuracy, a temperature impact prediction coefficient is calculated to serve as a temperature impact prediction scale.

6. The intelligent monitoring system for the electronic control unit of a pure electric heavy truck according to claim 5 is characterized in that: Temperature impact prediction module for: According to the historical temperature monitoring data of the power electronic control unit of the pure electric heavy truck, a sample temperature sequence set, a sample heavy truck feature information set, a sample road condition information set, and a sample predicted temperature set are collected; Divide the sample temperature sequence set, the sample heavy truck feature information set, the sample road condition information set and the sample predicted temperature set into K portions of temperature impact prediction supervision training data, where K is a positive integer; The K pieces of temperature impact prediction supervision training data are used to train to obtain K temperature impact prediction paths, and the temperature impact prediction channels are obtained by combining them; According to the temperature impact prediction scale and the K temperature impact prediction paths, at least one temperature impact prediction path is selected and calculated, the heavy truck feature information, road condition information and temperature sequence are respectively input, at least one predicted temperature is predicted, and the average is calculated to obtain the final predicted temperature.

7. The intelligent monitoring system for the electronic control unit of a pure electric heavy truck according to claim 1 is characterized in that: Monitoring result acquisition module, used for: Collecting a sample temperature set of the power electronic control unit, and marking different state levels for different sample temperatures to obtain a sample state parameter set, wherein the state level is negatively correlated with the sample temperature; Constructing a mapping relationship between the sample temperature set and the sample state parameter set to obtain a state monitoring classifier; The predicted temperature is input into the state monitoring classifier, and the state parameters of the power electronic control unit are obtained by mapping and classification as monitoring results.

8. A method for intelligent monitoring of an electronic control unit of a pure electric heavy truck, characterized in that: The method is performed by an intelligent monitoring system for an electronic control unit of a pure electric heavy truck according to any one of claims 1 to 7, comprising: Monitoring and obtaining a temperature sequence of a power electronic control unit of a pure electric heavy-duty truck, and collecting heavy-duty truck characteristic information of the pure electric heavy-duty truck; Obtain the image of the road ahead, identify the road condition, and obtain the road condition information and road condition accuracy; According to the road condition accuracy and the heavy truck characteristic information, a temperature impact prediction scale for temperature impact prediction is configured; according to the temperature impact prediction scale, the temperature impact prediction is performed according to the heavy truck characteristic information, the road condition information and the temperature sequence to obtain a predicted temperature; According to the predicted temperature, the state parameters of the power electronic control unit are obtained by classification as monitoring results.

Images