Commercial vehicle engine real working point identification and optimization method based on neural vector field alignment and knowledge base and related equipment
By combining a neural network vector field model based on remote monitoring data with a knowledge base, the actual operating point of commercial vehicle engines is identified and optimization suggestions are generated. This solves the problems of fuel consumption deviation and limited optimization effect in existing technologies, and achieves precise energy consumption optimization.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-14
- Publication Date
- 2026-03-13
AI Technical Summary
Existing technologies for analyzing and optimizing the energy consumption of commercial vehicle engines suffer from problems such as significant deviations between fuel consumption results and actual road conditions, distorted identification of operating points, and limited optimization effects, making it difficult to automatically output systematic optimization solutions.
Engine operating condition datasets are generated based on remote monitoring data from commercial vehicles, universal characteristic curves are plotted, real operating points are identified using a neural network vector field model, and optimization suggestions are generated by combining a knowledge base, including data cleaning, neural network training, and expert experience.
It achieves accurate identification of the operating point of commercial vehicle engines under real road conditions and quantifies optimization strategies, outputs reliable optimization suggestions, and improves the accuracy and systematic nature of energy consumption optimization.
Smart Images
Figure CN121659720A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of engine energy consumption analysis and optimization technology for commercial vehicles, and in particular to a method and related equipment for identifying and optimizing the real operating point of a commercial vehicle engine based on neural vector field alignment and knowledge base. Background Technology
[0002] Currently, technologies related to engine energy consumption analysis and optimization for commercial vehicles can be broadly categorized into two types: one relies on bench universal characteristic curves, which suffers from significant discrepancies between fuel consumption results and actual road conditions; the other directly utilizes remote monitoring data, but this approach often suffers from inaccurate operating point identification due to noise and incomplete cleaning of invalid points. Furthermore, both of these technologies can only qualitatively compare the differences between actual operating areas and economic zones, offering limited optimization effects on operating points and failing to automatically generate systematic optimization solutions for the R&D phase. Summary of the Invention
[0003] To address the aforementioned shortcomings, this application provides a method and related equipment for identifying and optimizing the true operating point of a commercial vehicle engine based on neural vector field alignment and a knowledge base. This method can accurately identify the engine's true operating point and output targeted optimization suggestions.
[0004] This application provides a method for identifying and optimizing the true operating point of a commercial vehicle engine according to a first aspect, the method comprising: An engine operating condition dataset is generated based on remote monitoring data from commercial vehicles. This data includes torque, speed, and contextual features at multiple operating points. A universal characteristic curve is plotted based on the engine operating condition dataset. The density distribution of torque and speed is statistically analyzed using this dataset, and the true operating points are identified based on this density distribution. A low-specific-fuel-consumption region is extracted from the universal characteristic curve as the economic zone, and the nearest offset vector from each true operating point to the boundary of the economic zone is calculated. A neural network vector field model is trained using the torque, speed, and contextual features of the true operating points as training samples and the nearest offset vector as a label. The torque, speed, and contextual features of the target operating point to be optimized are input into the trained neural network vector field model to obtain the offset direction from the target operating point to the economic zone. This offset direction is then matched with a knowledge base to generate optimization suggestions.
[0005] In some embodiments, generating an engine operating condition dataset based on remote monitoring data of commercial vehicles includes: calculating the engine power and specific fuel consumption corresponding to each operating point based on the torque, speed and context features corresponding to each operating point in the remote monitoring data to obtain an initial engine operating condition dataset; performing data cleaning on the initial engine operating condition dataset to obtain the engine operating condition dataset; the data cleaning includes low-pass filtering and effective value masking.
[0006] In some embodiments, plotting a universal characteristic curve based on an engine operating condition dataset includes: mapping torque and speed data in the engine operating condition dataset to a torque-speed plane, and plotting multiple contour maps of specific fuel consumption to form a universal characteristic curve.
[0007] In some embodiments, the neural network vector field model is trained using the torque, speed, and contextual features of the actual operating point as training samples and the nearest offset vector as a label. This includes: constructing a neural network vector field model, where the input of the neural network vector field model is the torque, speed, and contextual features of the operating point, and the output is the offset direction from the operating point to the economic zone; and training the neural network vector field model with alignment distance constraints, energy consumption monotonicity constraints, and feasible region constraints as training objectives to obtain a trained neural network vector field model.
[0008] In some embodiments, the alignment distance constraint is used to minimize the distance between the operating point and the economic zone after the offset direction adjustment; the energy consumption monotonicity constraint is used to ensure that the specific fuel consumption corresponding to the adjusted operating point is not higher than the specific fuel consumption corresponding to the operating point before adjustment; and the feasible region constraint is used to ensure that the adjusted operating point is within the physical boundaries of the engine's torque and speed.
[0009] In some embodiments, the knowledge base includes a mapping relationship between operating point deviation patterns and optimization measures obtained by mining and refining multi-vehicle data; matching the offset direction of the target operating point from the economic zone with the knowledge base to generate optimization suggestions, including: matching the offset direction of the target operating point from the economic zone with the operating point deviation patterns in the knowledge base; and invoking the corresponding optimization measures based on the matched operating point deviation patterns to generate optimization suggestions.
[0010] In some embodiments, the method further includes: when the offset direction from the target working point to the economic zone does not match the working point deviation pattern in the knowledge base, determining a new working point deviation pattern based on the offset direction; obtaining the optimization measures corresponding to the new working point deviation pattern confirmed by experts; and updating the knowledge base according to the new working point deviation pattern and its corresponding optimization measures.
[0011] According to a second aspect, this application provides a device for identifying and optimizing the true operating point of a commercial vehicle engine, the device comprising: The dataset generation module is used to generate engine operating condition datasets based on remote monitoring data from commercial vehicles; the remote monitoring data includes torque, speed, and contextual features at multiple operating points. The universal characteristic curve plotting module is used to plot universal characteristic curves based on engine operating condition datasets. The true operating point identification module is used to statistically analyze the density distribution of torque and speed based on the engine operating condition dataset, and to identify the true operating point based on the density distribution of torque and speed. The economic zone comparison module is used to extract the low specific fuel consumption area as the economic zone from the universal characteristic curve and calculate the nearest offset vector from each real operating point to the boundary of the economic zone. The neural vector field alignment module is used to train the neural network vector field model using torque, speed and context features of the real operating point as training samples and the nearest offset vector as the label. The optimization suggestion generation module is used to input the torque, speed and contextual features of the target operating point to be optimized into a trained neural network vector field model to obtain the offset direction of the target operating point from the economic zone. The offset direction of the target operating point from the economic zone is then matched with the knowledge base to generate optimization suggestions.
[0012] According to a third aspect, this application provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, it implements the steps of any of the commercial vehicle engine real operating point identification and optimization methods described in the above embodiments.
[0013] According to a fourth aspect, this application provides a computer device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the steps of any of the methods for identifying and optimizing the true operating point of a commercial vehicle engine in the above embodiments.
[0014] The embodiments described above in this application can accurately identify the operating point distribution of commercial vehicle engines under real road conditions, and learn the offset direction between the real operating point and the economic zone through a neural network vector field model, outputting quantifiable optimization strategies. A knowledge base is also established, which, by combining expert experience and data mining results, generates suggestions for engine development and control strategy optimization. Attached Figure Description
[0015] Figure 1 This is a flowchart of a method for identifying and optimizing the true operating point of a commercial vehicle engine according to one or more embodiments of this application; Figure 2 This is a schematic diagram of the structure of the commercial vehicle engine real operating point identification and optimization device in one or more embodiments of this application; Figure 3 This is a schematic diagram of the internal structure of a computer device according to one or more embodiments of this application. Detailed Implementation
[0016] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings. It should be understood that the described embodiments are merely some embodiments of this application, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.
[0017] In the following description, when referring to the accompanying drawings, the same numbers in different drawings denote the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.
[0018] In the description of this application, it should be understood that the terms "first," "second," "third," etc., are used only to distinguish similar objects and are not necessarily used to describe a specific order or sequence, nor should they be construed as indicating or implying relative importance. Those skilled in the art can understand the specific meaning of the above terms in this application according to the specific circumstances. Furthermore, in the description of this application, unless otherwise stated, "multiple" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. The character " / " generally indicates that the preceding and following related objects have an "or" relationship.
[0019] To address the shortcomings or deficiencies of related technologies, this application provides a method for identifying and optimizing the true operating point of a commercial vehicle engine. This method can accurately identify the distribution of the operating point of a commercial vehicle engine under real road conditions, and learn the offset direction between the true operating point and the economic zone through a neural network vector field model, outputting a quantifiable optimization strategy. Simultaneously, a knowledge base is established, combining expert experience and data mining results to generate suggestions for engine development and control strategy optimization.
[0020] In some exemplary embodiments of this application, such as Figure 1 As shown, the method includes steps S110 to S160, and each step is described in detail below.
[0021] S110: Generate engine operating condition dataset based on remote monitoring data of commercial vehicles.
[0022] Commercial vehicles are typically diesel or gas-powered vehicles with a gross vehicle weight of ≥3.5 tons used for transporting goods or passengers, including tractors, cargo trucks, and buses.
[0023] Remote monitoring data includes torque, speed, and contextual characteristics at multiple operating points, such as monitoring data for commercial vehicles over the past 100 hours. Specifically, remote monitoring data consists of CAN signals uploaded to the cloud platform by the vehicle via an onboard T-Box at fixed intervals. This data includes not only engine torque and speed but also load, gradient, instantaneous fuel consumption, vehicle speed, GPS altitude, atmospheric pressure, and coolant temperature. An operating point is the engine's operating state at a specific moment, which can be a two-dimensional coordinate system uniquely determined by torque and speed values. Contextual characteristics are vehicle and / or environmental variables, other than torque and speed, that have a quantifiable impact on engine fuel consumption. These can include vehicle weight, road gradient, instantaneous fuel consumption, ambient temperature, altitude, drag coefficient, and driver-demanded throttle opening.
[0024] The engine operating condition dataset is a structured collection of records generated based on remote monitoring data. Each record contains at least torque, speed, contextual features, and the corresponding instantaneous specific fuel consumption (BSFC).
[0025] In some embodiments, the step of generating an engine operating condition dataset based on remote monitoring data of commercial vehicles includes: calculating the engine power and specific fuel consumption corresponding to each operating point based on the torque, speed and context features corresponding to each operating point in the remote monitoring data to obtain an initial engine operating condition dataset; performing data cleaning on the initial engine operating condition dataset to obtain an engine operating condition dataset; the data cleaning includes low-pass filtering and effective value masking.
[0026] The engine power corresponding to the operating point can be calculated using the following formula: ; This refers to engine power. It refers to the engine's torque. This refers to the engine's angular velocity, which can be calculated using the following formula: =2πn / 60; n refers to the engine speed. The specific fuel consumption at the operating point is the ratio of the instantaneous fuel consumption at the operating point to the engine power.
[0027] The initial engine operating condition dataset refers to the structured record obtained after calculating engine power and specific fuel consumption from remote monitoring data. These records still contain noise and invalid points, and only physical quantity conversions have been performed. For the initial engine operating condition dataset, low-pass filtering and effective value masking can be applied sequentially. Low-pass filtering eliminates high-frequency noise in the initial engine operating condition dataset, while effective value masking removes invalid points such as idling speed and sudden changes, effectively reducing errors in subsequent real operating point identification operations. This eliminates the need to process invalid samples during subsequent model training, reducing computational load.
[0028] S120: Plot the universal characteristic curve based on the engine operating condition dataset.
[0029] The universal characteristic curve is a contour plot drawn with torque as the horizontal axis, speed as the vertical axis, and specific fuel consumption (BSFC) as the third dimension. It is used to characterize the fuel consumption distribution of an engine across the entire operating range.
[0030] The operation of drawing universal characteristic curves based on engine operating condition datasets can be done by first mapping the torque and speed data in the engine operating condition datasets to a torque-speed plane, and then drawing multiple contour maps of specific fuel consumption to form universal characteristic curves.
[0031] S130: Statistically analyze the torque-speed density distribution based on the engine operating condition dataset, and identify the true operating point based on the torque-speed density distribution.
[0032] The torque-speed density distribution is a probability density function obtained by performing a two-dimensional kernel density estimation on the torque-speed plane for all operating points in the engine operating condition dataset. It is used to quantify the frequency of occurrence in each region. The true operating points are a set of high-frequency operating points truncated according to the torque-speed density distribution and a cumulative probability threshold; these represent the most frequently used torque-speed combinations in actual road conditions. Specifically, all operating points in the engine operating condition dataset can be projected onto the torque-speed two-dimensional plane, and then a continuous density field can be generated using a standard two-dimensional kernel density estimation algorithm. Next, a truncation is performed according to a preset cumulative density ratio, such as 95%, retaining the highest frequency region as the main operating point interval, for example, the region of 1500 rpm and 850 Nm. The main operating point interval includes multiple true operating points.
[0033] S140: Extract the low specific fuel consumption region from the universal characteristic curve as the economic zone, and calculate the nearest offset vector from each real operating point to the boundary of the economic zone.
[0034] The economic zone is the connected region on the universal characteristic curve where the specific fuel consumption is less than the product of the preset minimum specific fuel consumption and a preset ratio value (such as 110%). In other words, it is the operating range where the engine's theoretical fuel consumption is optimal. For example, the region between 1300 and 1600 rpm and 700 and 900 Nm.
[0035] The closest offset vector from the actual operating point to the boundary of the economic zone refers to the shortest Euclidean distance vector pointing to the boundary of the economic zone in the torque-speed plane for that actual operating point. Its components represent the required torque increment adjustment. With speed increment .
[0036] S150: The neural network vector field model is trained using torque, speed and contextual features at the real operating point as training samples and the nearest offset vector as the label.
[0037] The neural network vector field model takes torque, speed, and contextual features as input and the nearest offset vector as the output label. It is a nonlinear mapping function trained using a fully connected or residual network structure. , It is used to predict the offset direction of any working point in a continuous domain.
[0038] The steps for training a neural network vector field model using torque, speed, and contextual features at the actual operating point as training samples and the nearest offset vector as the label include: constructing a neural network vector field model, where the input to the model is the torque, speed, and contextual features at the operating point, and the output is the offset direction from the operating point to the economic zone; and training the model with alignment distance constraints, energy consumption monotonicity constraints, and feasible region constraints as training objectives to obtain a trained neural network vector field model.
[0039] The trained neural network vector field can be visualized as an adjustment arrow diagram, intuitively showing the optimization direction. Among them, the alignment distance constraint is used to minimize the distance between the operating point and the economic zone after the offset direction adjustment; the energy consumption monotonicity constraint is used to ensure that the specific fuel consumption corresponding to the adjusted operating point is not higher than the specific fuel consumption corresponding to the operating point before adjustment; and the feasible region constraint is used to ensure that the adjusted operating point is within the physical boundaries of the engine's torque and speed.
[0040] Alignment distance constraints ensure that the arrow points to the economic zone, energy consumption monotonicity constraints ensure that the adjustment is more fuel-efficient than the original, and feasible region constraints lock the torque and speed within physical limits. This embodiment combines the above three constraints to make the offset direction given by the neural network vector field model both fuel-efficient and directly applicable, without the need for post-processing verification.
[0041] S160: Input the torque, speed and contextual features of the target operating point to be optimized into the trained neural network vector field model to obtain the offset direction of the target operating point from the economic zone. Match the offset direction of the target operating point from the economic zone with the knowledge base to generate optimization suggestions.
[0042] The target operating point is the engine operating condition to be optimized that has not yet fallen into the economic zone, and its torque, speed and contextual features are used as inputs to the trained model.
[0043] The knowledge base can specifically be a database. It includes expert experience rules, such as optimization rules summarized by industry experts, like "excessive dwell time in the high-speed zone → optimize shift logic," and "excessive proportion of low-speed, high-torque → improve turbocharging and fuel injection strategies." The knowledge base also includes data-driven knowledge, such as the mapping relationship between operating point deviation patterns and optimization measures obtained by mining and refining multi-vehicle data. Multi-vehicle data includes historical remote monitoring data and operating point optimization measures for multiple commercial vehicles. The operating point deviation pattern (or simply deviation pattern) refers to a template abstracted from the common offset characteristics between the actual operating point and the economic zone. It can be represented by a triplet of torque-speed range, offset vector direction and length, and contextual features (such as heavy load, hill climbing, etc.). This is used to classify similar high-fuel-consumption scenarios and directly associate them with one or more standard optimization measures.
[0044] In some embodiments, the offset direction from the target working point to the economic zone is matched with the knowledge base to generate optimization suggestions, including: matching the offset direction from the target working point to the economic zone with the working point deviation patterns in the knowledge base; and invoking corresponding optimization measures based on the matched working point deviation patterns to generate optimization suggestions.
[0045] For example, the offset direction of the model output can be... The vector is then vectorized, and the vector in the offset direction is compared with the pre-set deviation pattern vector in the knowledge base using cosine similarity. Next, the deviation pattern that is most similar to the vector and has a similarity greater than a preset similarity threshold (such as 0.85) is extracted from the knowledge base as the matching deviation pattern. Finally, the optimization measures bound to the deviation pattern are directly called as optimization suggestions for output.
[0046] For example, suppose that the main operating point of a commercial vehicle is concentrated in the 1500rpm and 850Nm range, and its economic zone is the 1300 to 1600rpm and 700 to 900Nm range. By comparing the actual operating point and the economic zone, it is found that the overlap is less than 60%. After inputting the characteristics of the target operating point (i.e., torque, speed and context features) into the neural network vector field model, the model outputs "reduce the speed by about 300rpm in the high speed range and reduce the torque by 50Nm in the high torque range". The offset direction of the model output is matched with the knowledge base to generate the following suggestions: (1) Control strategy: optimize the shift logic and shift the gear earlier; (2) Engine development: improve the combustion efficiency in the low speed range and improve the turbocharger response; (3) Vehicle matching: optimize the transmission ratio and reduce high speed operation.
[0047] In some embodiments, the method further includes: when the offset direction from the target working point to the economic zone does not match the working point deviation pattern in the knowledge base, determining a new working point deviation pattern based on the offset direction; obtaining the optimization measures corresponding to the new working point deviation pattern confirmed by experts; and updating the knowledge base according to the new working point deviation pattern and its corresponding optimization measures.
[0048] The offset direction from the target operating point to the economic zone does not match the operating point deviation patterns in the knowledge base. This could mean that the vector similarity between the offset direction and each deviation pattern is lower than a preset similarity threshold (e.g., 0.85). If the offset direction from the target operating point to the economic zone does not match any of the operating point deviation patterns in the knowledge base, the offset direction is automatically registered as a new deviation pattern and temporarily stored in the candidate area. The backend can push the corresponding torque-speed range and suggested adjustment amount to the expert terminal. After expert confirmation or revision, it is bound to the formal optimization measures. Once approved, it is immediately written into the knowledge base and the model matching list is incrementally updated, thereby realizing the dynamic updating of the knowledge base.
[0049] It should be noted that, regarding the steps included in the commercial vehicle engine true operating point identification and optimization method provided in any of the above embodiments, unless explicitly stated herein, there is no strict order restriction on the execution of these steps; they can be executed in other orders. Furthermore, at least some of these steps may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least a portion of the sub-steps or stages of other steps.
[0050] Based on the same inventive concept, this application also provides a device for identifying and optimizing the true operating point of a commercial vehicle engine. In some embodiments, such as Figure 2 As shown, the commercial vehicle engine real operating point identification and optimization device includes the following modules: The dataset generation module 110 is used to generate an engine operating condition dataset based on remote monitoring data of commercial vehicles; the remote monitoring data includes torque, speed and contextual features at multiple operating points; The universal characteristic curve plotting module 120 is used to plot universal characteristic curves based on engine operating condition datasets. The real operating point identification module 130 is used to statistically analyze the density distribution of torque and speed based on the engine operating condition dataset, and to identify the real operating point based on the density distribution of torque and speed. The economic zone comparison module 140 is used to extract the low specific fuel consumption area from the universal characteristic curve as the economic zone and calculate the nearest offset vector from each real operating point to the boundary of the economic zone. The neural vector field alignment module 150 is used to train the neural network vector field model using the torque, speed and context features of the real operating point as training samples and the nearest offset vector as the label. The optimization suggestion generation module 160 is used to input the torque, speed and context features of the target operating point to be optimized into the trained neural network vector field model to obtain the offset direction of the target operating point from the economic zone, and match the offset direction of the target operating point from the economic zone with the knowledge base to generate optimization suggestions.
[0051] In some embodiments, the dataset generation module 110 is used to calculate the engine power and specific fuel consumption corresponding to each operating point based on the torque, speed and context features corresponding to each operating point in the remote monitoring data, to obtain the initial engine operating condition dataset; and to perform data cleaning on the initial engine operating condition dataset to obtain the engine operating condition dataset; the data cleaning includes low-pass filtering and effective value masking.
[0052] In some embodiments, the universal characteristic curve plotting module 120 is used to map the torque and speed data in the engine operating condition dataset to the torque-speed plane and plot multiple contour lines of specific fuel consumption to form a universal characteristic curve.
[0053] In some embodiments, the neural vector field alignment module 150 is used to construct a neural network vector field model. The input of the neural network vector field model is the torque, speed and context features of the operating point, and the output is the offset direction from the operating point to the economic zone. The neural network vector field model is trained with alignment distance constraints, energy consumption monotonicity constraints and feasible region constraints as training objectives to obtain a trained neural network vector field model.
[0054] In some embodiments, the alignment distance constraint is used to minimize the distance between the operating point and the economic zone after the offset direction adjustment; the energy consumption monotonicity constraint is used to ensure that the specific fuel consumption corresponding to the adjusted operating point is not higher than the specific fuel consumption corresponding to the operating point before adjustment; and the feasible region constraint is used to ensure that the adjusted operating point is within the physical boundaries of the engine's torque and speed.
[0055] In some embodiments, the knowledge base includes a mapping relationship between operating point deviation patterns and optimization measures obtained by mining and refining multi-vehicle data; an optimization suggestion generation module 160 is used to match the offset direction of the target operating point from the economic zone with the operating point deviation patterns in the knowledge base; and to call the corresponding optimization measures based on the matched operating point deviation patterns to generate optimization suggestions.
[0056] In some embodiments, the apparatus further includes a knowledge base update module. The knowledge base update module is configured to, when the offset direction from the target operating point to the economic zone does not match the operating point deviation pattern in the knowledge base, determine a new operating point deviation pattern based on the offset direction; obtain optimization measures corresponding to the new operating point deviation pattern confirmed by experts; and update the knowledge base according to the new operating point deviation pattern and its corresponding optimization measures.
[0057] Specific limitations regarding the device for identifying and optimizing the true operating point of commercial vehicle engines can be found in the limitations of the method for identifying and optimizing the true operating point of commercial vehicle engines described above, and will not be repeated here. Each module in the aforementioned device for identifying and optimizing the true operating point of commercial vehicle engines can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.
[0058] This application also provides a computer device. In some embodiments, the computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it can implement the steps of the commercial vehicle engine real operating point identification and optimization method provided in any of the above embodiments.
[0059] In some embodiments, the internal structure diagram of a computer device may be as follows: Figure 3 As shown, the computer device includes a processor, memory, and network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database stores remote monitoring data and knowledge bases for commercial vehicles; the specific stored data may also be as defined in the above method embodiments. The network interface communicates with external terminals via a network connection. When executed by the processor, the computer program implements a method for identifying and optimizing the true operating point of a commercial vehicle engine.
[0060] Those skilled in the art will understand that Figure 3 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.
[0061] This application also provides a computer-readable storage medium, in some embodiments of which a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, it implements the steps of the commercial vehicle engine real operating point identification and optimization method provided in any of the above embodiments.
[0062] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.
[0063] Those skilled in the art will understand that implementing all or part of the processes in the above method embodiments can be accomplished by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), Synchlink, DRAM (SLDRAM), memory bus, direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.
[0064] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0065] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.
Claims
1. A method for identifying and optimizing the true operating point of a commercial vehicle engine, characterized in that, The method includes: Engine operating condition dataset is generated based on remote monitoring data from commercial vehicles; the remote monitoring data includes torque, speed, and contextual features at multiple operating points. Plot the universal characteristic curve based on the engine operating condition dataset; The density distribution of torque and speed is statistically analyzed based on the engine operating condition dataset, and the actual operating point is identified based on the density distribution of torque and speed. The low specific fuel consumption region is extracted from the universal characteristic curve as the economic zone, and the nearest offset vector from each actual operating point to the boundary of the economic zone is calculated. The neural network vector field model is trained using the torque, speed, and contextual features of the actual operating point as training samples and the nearest offset vector as the label. The torque, speed, and contextual features of the target operating point to be optimized are input into the trained neural network vector field model to obtain the offset direction of the target operating point from the economic zone. The offset direction of the target operating point from the economic zone is then matched with the knowledge base to generate optimization suggestions.
2. The method as described in claim 1, characterized in that, The engine operating condition dataset generated from the remote monitoring data of commercial vehicles includes: Based on the torque, speed and context features corresponding to each operating point in the remote monitoring data, the engine power and specific fuel consumption corresponding to each operating point are calculated to obtain the initial dataset of engine operating conditions. The initial dataset of engine operating conditions is cleaned to obtain the engine operating condition dataset; the data cleaning includes low-pass filtering and effective value masking.
3. The method as described in claim 1, characterized in that, The step of plotting the universal characteristic curve based on the engine operating condition dataset includes: The torque and speed data in the engine operating condition dataset are mapped to the torque-speed plane, and multiple contour maps of specific fuel consumption are drawn to form a universal characteristic curve.
4. The method as described in claim 1, characterized in that, The step of training the neural network vector field model using the torque, speed, and contextual features of the actual operating point as training samples and the nearest offset vector as the label includes: A neural network vector field model is constructed. The input of the neural network vector field model is the torque, speed and context features of the operating point, and the output is the offset direction of the operating point from the economic zone. The neural network vector field model is trained using alignment distance constraints, energy consumption monotonicity constraints, and feasible region constraints as training objectives to obtain a trained neural network vector field model.
5. The method as described in claim 4, characterized in that, The alignment distance constraint is used to minimize the distance between the operating point and the economic zone after the offset direction adjustment; the energy consumption monotonicity constraint is used to ensure that the specific fuel consumption corresponding to the adjusted operating point is not higher than the specific fuel consumption corresponding to the operating point before adjustment; the feasible region constraint is used to ensure that the adjusted operating point is within the physical boundaries of the engine's torque and speed.
6. The method as described in claim 1, characterized in that, The knowledge base includes the mapping relationship between working point deviation patterns and optimization measures obtained by mining and refining multi-vehicle data; The step of matching the offset direction of the target working point from the economic zone with the knowledge base to generate optimization suggestions includes: The offset direction of the target working point from the economic zone is matched with the working point deviation pattern in the knowledge base; Based on the matched operating point deviation pattern, the corresponding optimization measures are invoked to generate the optimization suggestions.
7. The method as described in claim 6, characterized in that, The method further includes: When the offset direction of the target working point from the economic zone does not match the working point deviation pattern in the knowledge base, a new working point deviation pattern is determined based on the offset direction. Obtain the optimization measures corresponding to the new operating point deviation pattern confirmed by experts, and update the knowledge base according to the new operating point deviation pattern and its corresponding optimization measures.
8. A device for identifying and optimizing the true operating point of a commercial vehicle engine, characterized in that, The device includes: The dataset generation module is used to generate engine operating condition datasets based on remote monitoring data of commercial vehicles; the remote monitoring data includes torque, speed and contextual features at multiple operating points; The universal characteristic curve plotting module is used to plot universal characteristic curves based on the engine operating condition dataset. The true operating point identification module is used to statistically analyze the density distribution of torque and speed based on the engine operating condition dataset, and to identify the true operating point based on the density distribution of torque and speed. The economic zone comparison module is used to extract the low specific fuel consumption area as the economic zone from the universal characteristic curve, and calculate the nearest offset vector from each of the real operating points to the boundary of the economic zone. The neural vector field alignment module is used to train the neural network vector field model using the torque, speed and context features of the real operating point as training samples and the nearest offset vector as a label. The optimization suggestion generation module is used to input the torque, speed and contextual features of the target operating point to be optimized into the trained neural network vector field model, obtain the offset direction of the target operating point from the economic zone, match the offset direction of the target operating point from the economic zone with the knowledge base, and generate optimization suggestions.
9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.
10. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.