A method and system for multi-objective optimization of engine operating points for a hybrid vehicle

Abstract

The application discloses a method and system for multi-objective optimization of engine operating points of a hybrid vehicle. The method comprises: obtaining engine operating data of a target engine at multiple steady-state operating points; determining the effective fuel consumption rate and multiple effective pollutant emission indicators corresponding to each steady-state operating point, and determining a candidate operating point set according to the effective fuel consumption rate; determining the emission correlation coefficient between each effective pollutant emission indicator and the effective fuel consumption rate; determining the emission weight in a manner that the lower the correlation degree, the higher the emission weight; constructing a multi-objective optimization function according to the effective fuel consumption rate, the effective pollutant emission indicator, the emission weight and the target power deviation evaluation information, and performing operating point optimization on multiple target power requirements to generate an engine optimized operating point set. Thus, the subjectivity of emission weight setting can be reduced, and fuel economy and multi-pollutant emission control can be considered under the target power constraint.

Description

Technical Field

[0001] This application relates to the field of hybrid electric vehicle technology, and in particular to a multi-objective optimization method and system for engine operating point in hybrid vehicle models. Background Technology

[0002] With the development of hybrid electric vehicle (HEV) technology, vehicle powertrain systems not only need to meet power demands under different operating conditions, but also need to consider fuel economy, emission control, and overall vehicle control efficiency. As a crucial power source in HEVs, the engine's operating state is typically determined by both speed and torque; different engine operating points correspond to varying levels of fuel consumption and emissions. Therefore, in the energy management process of HEVs, rationally determining the engine's operating point under different power demands is a critical factor affecting overall vehicle fuel consumption, emissions, and powertrain control performance.

[0003] In existing technologies, engine operating points are typically selected based on engine universal characteristic diagrams, empirical calibration data, or preset optimization targets. For example, some solutions determine the engine operating range or operating line with low fuel consumption as the primary objective to improve vehicle fuel economy; others consider fuel consumption and pollutant emissions as multiple evaluation indicators to achieve a relatively balanced engine operating result. However, the engine combustion process is complex, and fuel consumption and various pollutant emissions do not always show a consistent trend under different operating conditions. A single operating point selection method that prioritizes fuel economy may lead to a deterioration of some emission indicators. Furthermore, when using a multi-indicator comprehensive evaluation method, the importance of each evaluation indicator usually relies on manual experience, making it difficult to accurately reflect the differences in fuel consumption and emission characteristics under different engines and operating conditions, resulting in insufficient stability and applicability of the operating point optimization results. In addition, during the actual operation of hybrid vehicles, the engine operating point also needs to be matched with the vehicle's power requirements, battery status, and drive system control strategy. Without a foundation of candidate operating points for vehicle control, the computational burden on the control strategy can easily increase, affecting the overall optimization effect of fuel consumption and emissions.

[0004] Therefore, in the process of optimizing the operating point of hybrid vehicle engines, it is difficult to balance fuel economy and emissions of multiple pollutants, the determination of engine operating point evaluation parameters relies on experience and lacks objectivity, and the optimized operating point has poor compatibility with the vehicle's energy management strategy, which are urgent problems to be solved. Summary of the Invention

[0005] This application provides a multi-objective optimization method and system for engine operating point in hybrid vehicles, aiming to solve the problems in the existing technology of optimizing the engine operating point of hybrid electric vehicles, such as the difficulty in synergistically balancing fuel economy and emissions of multiple pollutants, the dependence of engine operating point evaluation parameters on experience and insufficient objectivity, and the poor adaptability of the optimized operating point to the vehicle's energy management strategy.

[0006] Firstly, a multi-objective optimization method for engine operating point in hybrid vehicles, the method comprising: The engine operating condition data of the target engine at multiple steady-state operating points are obtained. The engine operating condition data includes engine speed, engine torque, fuel consumption data and various pollutant emission data. Based on the engine operating data, the effective fuel consumption rate and multiple effective pollutant emission indicators corresponding to each steady-state operating point are determined, and a candidate operating point set is determined from multiple steady-state operating points based on the effective fuel consumption rate and preset economic screening conditions. Based on the effective fuel consumption rate corresponding to each candidate operating point in the candidate operating point set and multiple effective pollutant emission indicators, determine the emission correlation coefficient between each effective pollutant emission indicator and the effective fuel consumption rate; The emission weights corresponding to each effective pollutant emission index are determined based on the emission correlation coefficients; wherein, the lower the correlation between the effective pollutant emission index and the effective fuel consumption rate, the higher the emission weight corresponding to the effective pollutant emission index. Based on the effective fuel consumption rate, multiple effective pollutant emission indicators, emission weights, and target power deviation evaluation information, a multi-objective optimization function for engine operating point is constructed. Based on the multi-objective optimization function of the engine operating point, the engine operating point is optimized for multiple target power requirements in the candidate operating point set to obtain the target engine operating point corresponding to each target power requirement; wherein, the target engine operating point includes the target engine speed and the target engine torque. Based on the target engine operating points corresponding to the multiple target power requirements, an optimized engine operating point set is generated. This optimized engine operating point set is used by the hybrid vehicle energy management strategy to select the target engine operating point based on real-time power requirements.

[0007] Optionally, in the above scheme, the engine operating condition data may also include exhaust flow data; The step of determining the effective fuel consumption rate and multiple effective pollutant emission indicators corresponding to each steady-state operating point based on the engine operating data includes: The engine output power corresponding to each steady-state operating point is determined based on the engine speed and engine torque corresponding to each steady-state operating point. Based on the fuel consumption data and engine output power corresponding to each steady-state operating point, determine the effective fuel consumption rate corresponding to each steady-state operating point; Based on the emission data of various pollutants, exhaust flow data and engine output power corresponding to each steady-state operating point, determine multiple effective pollutant emission indicators corresponding to each steady-state operating point. Based on the effective fuel consumption rate and multiple effective pollutant emission indicators, operating condition index data for characterizing engine fuel consumption and emission characteristics are obtained.

[0008] Optionally, in the above scheme, the emission data of multiple pollutants includes total hydrocarbon emission data, carbon monoxide emission data, and nitrogen oxide emission data; The step of determining multiple effective pollutant emission indicators corresponding to each steady-state operating point based on various pollutant emission data, exhaust flow data, and engine output power includes: The effective total hydrocarbon emission index is determined based on the total hydrocarbon emission data, the exhaust flow data, and the corresponding engine output power. The effective carbon monoxide emission index is determined based on the carbon monoxide emission data, the exhaust flow data, and the corresponding engine output power. The effective nitrogen oxide emission index is determined based on the nitrogen oxide emission data, the exhaust flow data, and the corresponding engine output power. Based on the effective total hydrocarbon emission index, the effective carbon monoxide emission index, and the effective nitrogen oxide emission index, multiple effective pollutant emission indices are obtained.

[0009] Optionally, in the above scheme, determining the candidate operating point set from multiple steady-state operating points based on the effective fuel consumption rate and preset economic screening conditions includes: The effective fuel consumption rate corresponding to each steady-state operating point is compared with the preset fuel consumption rate threshold to obtain the economic judgment result corresponding to each steady-state operating point. Based on the economic assessment results, steady-state operating points that do not meet the preset economic screening conditions are removed from the multiple steady-state operating points to obtain the set of operating points in the economic zone. The candidate operating point set is determined based on the engine speed, engine torque, effective fuel consumption rate, and multiple effective pollutant emission indicators corresponding to each operating point in the economic zone operating point set.

[0010] Optionally, in the above scheme, determining the emission correlation coefficient between each effective pollutant emission index and the effective fuel consumption rate based on the effective fuel consumption rate corresponding to each candidate operating point in the candidate operating point set and multiple effective pollutant emission indices includes: The effective fuel consumption rate corresponding to each candidate operating point in the candidate operating point set is normalized to obtain normalized fuel consumption data. The multiple effective pollutant emission indicators corresponding to each candidate working point in the candidate working point set are normalized to obtain multiple normalized emission indicator data. Based on the normalized fuel consumption data and the normalized emission index data, the linear correlation parameters between each effective pollutant emission index and the effective fuel consumption rate are calculated to obtain multiple emission correlation coefficients.

[0011] Optionally, in the above scheme, determining the emission weight corresponding to each effective pollutant emission indicator based on the emission correlation coefficient includes: The degree of correlation between each effective pollutant emission index and the effective fuel consumption rate is determined based on the absolute value of each emission correlation coefficient. Based on the correlation degree of the multiple effective pollutant emission indicators, determine the correlation degree ranking information of the multiple effective pollutant emission indicators; Based on the correlation ranking information and the preset reverse weight allocation rule, the initial emission weights corresponding to multiple effective pollutant emission indicators are determined; The initial emission weights are normalized to obtain the emission weights corresponding to each effective pollutant emission index; wherein, the emission weights corresponding to effective pollutant emission indices with lower correlation are higher.

[0012] Optionally, in the above scheme, the step of constructing a multi-objective optimization function for the engine operating point based on the effective fuel consumption rate, multiple effective pollutant emission indicators, the emission weights, and target power deviation evaluation information includes: Based on the effective fuel consumption rate, determine the fuel economy evaluation items; Based on the multiple effective pollutant emission indicators and their corresponding emission weights, emission evaluation items are determined; Based on the deviation between the engine output power corresponding to the candidate operating point and the target power requirement, target power deviation evaluation information is determined to constrain the deviation of the candidate operating point from the target power requirement. Based on the fuel economy evaluation item, the emission evaluation item, and the target power deviation evaluation information, a multi-objective optimization function for the engine operating point is constructed.

[0013] Optionally, in the above scheme, the step of optimizing the engine operating point for multiple target power requirements based on the engine operating point multi-objective optimization function to obtain the target engine operating point corresponding to each target power requirement includes: Based on the engine speed, engine torque, effective fuel consumption rate, multiple effective pollutant emission indicators, and engine output power corresponding to each candidate operating point in the candidate operating point set, engine operating point search data is constructed. For any target power requirement, based on the engine operating point multi-objective optimization function and the preset global optimization algorithm, the candidate operating points in the engine operating point search data are iteratively searched to obtain the candidate optimal operating point in the iterative process. If the candidate optimal operating point satisfies the preset convergence condition, the candidate optimal operating point is determined as the target engine operating point corresponding to the target power requirement; By iterating through multiple target power requirements, the target engine operating point corresponding to each target power requirement is obtained.

[0014] Optionally, in the above scheme, the engine optimized operating point set is used for the hybrid vehicle energy management strategy to select the target engine operating point according to real-time power demand, including: Obtain real-time operating status information of hybrid vehicles, including real-time power demand and power battery state of charge; Based on the state of charge of the power battery, determine whether the hybrid vehicle is in the power maintenance phase, and obtain the energy management phase judgment result; When the energy management phase judgment result indicates that the hybrid vehicle is in the battery maintenance phase, a target engine operating point whose target power demand meets the preset matching condition with the real-time power demand is selected from the set of optimized engine operating points based on the real-time power demand. Based on the matched target engine operating point, a target engine speed command and a target engine torque command are generated so that the engine of the hybrid vehicle operates in accordance with the target engine speed command and the target engine torque command.

[0015] Secondly, a multi-objective optimization system for engine operating point in hybrid vehicles, the system comprising: The data acquisition module is used to acquire engine operating condition data of the target engine at multiple steady-state operating points. The engine operating condition data includes engine speed, engine torque, fuel consumption data, and various pollutant emission data. The indicator determination module is used to determine the effective fuel consumption rate and multiple effective pollutant emission indicators corresponding to each of the engine operating condition points based on the engine operating condition data, and to determine a set of candidate operating points from the multiple steady-state operating points based on the effective fuel consumption rate and preset economic screening conditions. The correlation determination module is used to determine the emission correlation coefficient between each effective pollutant emission index and the effective fuel consumption rate based on the effective fuel consumption rate corresponding to each candidate operating point in the candidate operating point set and multiple effective pollutant emission indicators. The weight determination module is used to determine the emission weight corresponding to each effective pollutant emission index based on the emission correlation coefficient; wherein, the lower the correlation between the effective pollutant emission index and the effective fuel consumption rate, the higher the emission weight corresponding to the effective pollutant emission index. The function construction module is used to construct a multi-objective optimization function for the engine operating point based on the effective fuel consumption rate, multiple effective pollutant emission indicators, the emission weights, and target power deviation evaluation information. The operating point optimization module is used to optimize the engine operating point for multiple target power requirements in the candidate operating point set based on the engine operating point multi-objective optimization function, and obtain the target engine operating point corresponding to each target power requirement; wherein, the target engine operating point includes the target engine speed and the target engine torque; The operating point set generation module is used to generate an optimized engine operating point set based on the target engine operating points corresponding to the multiple target power requirements. The optimized engine operating point set is used for the hybrid vehicle energy management strategy to select the target engine operating point according to the real-time power requirements.

[0016] Compared with the prior art, this application has at least the following beneficial effects: This application, based on further analysis and research of existing technical problems, recognizes that existing technologies in the optimization of engine operating points for hybrid electric vehicles suffer from several issues: difficulty in synergistically considering fuel economy and multi-pollutant emissions; reliance on experience and insufficient objectivity in determining engine operating point evaluation parameters; and poor adaptability between the optimized operating point and the vehicle's energy management strategy. This application addresses these issues by acquiring engine operating condition data for the target engine at multiple steady-state operating points, and using this data to determine the effective fuel consumption rate and multiple effective pollutant emission indicators corresponding to each steady-state operating point. This allows for the evaluation of fuel economy and multi-pollutant emission levels at the engine operating point on the same operating condition basis. Furthermore, by determining a candidate operating point set based on the effective fuel consumption rate, subsequent optimization can be limited to the operating condition range where fuel economy requirements are met, preventing the optimization results from deviating from the engine's economical operating range. Based on this, the application determines the emission correlation coefficient between each effective pollutant emission indicator and the effective fuel consumption rate, and assigns higher emission weights to each effective pollutant emission indicator according to the principle that lower correlation values ​​result in higher emission weights. Emission weights are introduced to provide stronger constraints on pollutant emission indicators that are weakly correlated with effective fuel consumption rate or difficult to improve synchronously with fuel consumption optimization, thereby reducing the reliance on human experience in setting emission evaluation parameters. Subsequently, a multi-objective optimization function for engine operating points is constructed based on effective fuel consumption rate, multiple effective pollutant emission indicators, emission weights, and target power deviation evaluation information. Engine operating points are then optimized for multiple target power requirements within the candidate operating point set. This allows for the acquisition of target engine speeds and torques that balance fuel economy and multi-pollutant emissions under the constraint of target power requirements. Finally, an optimized engine operating point set is generated based on the target engine operating points corresponding to multiple target power requirements. This enables hybrid vehicle energy management strategies to directly select target engine operating points based on real-time power requirements. This addresses the issues of difficulty in synergistically balancing fuel economy and multi-pollutant emissions, insufficient objectivity in determining engine operating point evaluation parameters, and poor compatibility between optimized operating points and vehicle energy management strategies during hybrid vehicle engine operating point optimization. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating a multi-objective optimization method for engine operating point in a hybrid vehicle according to an embodiment of this application. Figure 2 This application provides an embodiment of an engine's effective fuel consumption and pollutant emission characteristics diagram; wherein, Figure 2 In this context, (a) is used to represent the effective fuel consumption rate characteristic. Figure 2 (b) in the text is used to represent the effective carbon monoxide emission characteristics. Figure 2 (c) in the text is used to represent the effective nitrogen oxide emission characteristics. Figure 2(d) in the figure is used to represent the effective total hydrocarbon emission characteristics; Figure 3 A correlation analysis diagram between effective fuel consumption rate and multiple effective pollutant emission indicators provided in one embodiment of this application; Figure 4 A flowchart for optimizing the engine operating point is provided for one embodiment of this application. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0019] In one embodiment, such as Figure 1 As shown, a multi-objective optimization method for engine operating points in hybrid vehicles is provided. This method can be applied to offline optimization scenarios of hybrid vehicle engine operating points, and the optimized engine operating point set can also be used as the basis for candidate operating points in the energy management strategy of hybrid vehicles. The target engine can be a dedicated hybrid gasoline engine, such as a 2.0L naturally aspirated hybrid gasoline engine; the hybrid vehicle can be a power-split hybrid vehicle, or other hybrid vehicles that can control engine operation based on target engine speed and target torque.

[0020] In practical implementation, the first step is to acquire engine operating condition data for the target engine at multiple steady-state operating points. These steady-state operating points can be obtained through engine bench testing, for example, by conducting steady-state performance tests on an engine bench according to preset speed and load ranges. The preset speed range can cover the target engine's commonly used operating range, such as 1000 r / min to 5600 r / min; the preset load range can cover 0 to 100% throttle opening. At each steady-state operating point, engine speed, engine torque, fuel consumption data, and various pollutant emission data are simultaneously collected. Fuel consumption data can be collected using a transient fuel flow meter, and pollutant emission data can be collected using an exhaust gas analyzer. The various pollutant emission data can include total hydrocarbon emissions, carbon monoxide emissions, and nitrogen oxide emissions.

[0021] Then, based on engine operating condition data, the effective fuel consumption rate and multiple effective pollutant emission indicators corresponding to each steady-state operating point are determined. The effective fuel consumption rate represents the fuel consumption level corresponding to a unit of output power of the target engine; the multiple effective pollutant emission indicators represent the different pollutant emission levels corresponding to a unit of output power of the target engine. Since raw pollutant emission data is usually collected in concentration form, it can be combined with exhaust flow data and engine output power to convert pollutant concentration data into effective pollutant emission indicators with the same dimensions as or comparable to the effective fuel consumption rate. This yields the following... Figure 2 The data shown are the engine's effective fuel consumption and pollutant emission characteristics. Figure 2 The effective fuel consumption rate characteristics, effective carbon monoxide emission characteristics, effective nitrogen oxide emission characteristics, and effective total hydrocarbon emission characteristics can be used to characterize the fuel consumption and emission distribution under different engine speeds and engine torque operating points.

[0022] Furthermore, based on the effective fuel consumption rate and preset economic screening criteria, a set of candidate operating points is determined from multiple steady-state operating conditions. The preset economic screening criteria can be that the effective fuel consumption rate is no greater than a preset fuel consumption rate threshold, for example, an effective fuel consumption rate no greater than 250 g / kWh. This screening process eliminates operating points in uneconomical zones with poor fuel economy, retaining only steady-state operating points with relatively good fuel economy as candidate operating points, thus avoiding subsequent optimization processes being conducted in clearly uneconomical operating areas.

[0023] Then, based on the effective fuel consumption rate and multiple effective pollutant emission indicators corresponding to each candidate operating point in the candidate operating point set, the emission correlation coefficient between each effective pollutant emission indicator and the effective fuel consumption rate is determined. Specifically, the effective fuel consumption rate and multiple effective pollutant emission indicators can be normalized first, and then the linear correlation parameters between each effective pollutant emission indicator and the effective fuel consumption rate can be calculated based on the normalized data. For example... Figure 3 As shown, the correlations between effective total hydrocarbon emissions, effective carbon monoxide emissions, and effective nitrogen oxide emissions and effective fuel consumption rate can be obtained. In a specific example, effective carbon monoxide emissions are strongly positively correlated with effective fuel consumption rate, effective total hydrocarbon emissions are weakly positively correlated with effective fuel consumption rate, and effective nitrogen oxide emissions are approximately nonlinearly correlated with effective fuel consumption rate.

[0024] Subsequently, the emission weights corresponding to each effective pollutant emission indicator are determined based on the emission correlation coefficient. Specifically, the lower the correlation between an effective pollutant emission indicator and the effective fuel consumption rate, the higher the emission weight of that effective pollutant emission indicator. In other words, for pollutant emission indicators that show a strong synchronous change with the effective fuel consumption rate, their weight in the emission assessment can be relatively reduced; for pollutant emission indicators that are weakly correlated with or essentially unrelated to the effective fuel consumption rate, their weight in the emission assessment is increased, thus providing stronger constraints on these pollutant emission indicators during the optimization process at the operating point.

[0025] Then, based on the effective fuel consumption rate, multiple effective pollutant emission indicators, emission weights, and target power deviation evaluation information, a multi-objective optimization function for engine operating points is constructed. This multi-objective optimization function can include a fuel economy evaluation term, an emission evaluation term, and a target power deviation evaluation term. The fuel economy evaluation term evaluates the fuel consumption level of candidate operating points, the emission evaluation term evaluates the overall pollutant emission level of candidate operating points, and the target power deviation evaluation term constrains the degree to which the engine output power corresponding to the candidate operating point deviates from the target power requirement. Therefore, the multi-objective optimization function for engine operating points can comprehensively evaluate fuel consumption and emissions while meeting the target power requirement.

[0026] Furthermore, based on the multi-objective optimization function for engine operating points, engine operating points are optimized for multiple target power demands within the candidate operating point set to obtain the target engine operating point corresponding to each target power demand. These multiple target power demands can cover the common engine power demand range for hybrid vehicles during the battery maintenance phase or other engine-driven phases, such as multiple target power points within the range of 10kW to 70kW. For any target power demand, the candidate operating point with the best comprehensive evaluation result can be searched within the candidate operating point set and determined as the target engine operating point corresponding to that target power demand. The target engine operating point includes the target engine speed and the target engine torque.

[0027] Finally, an optimized engine operating point set is generated based on the target engine operating points corresponding to multiple target power demands. This optimized engine operating point set can be stored as a table of power, speed, and torque correspondences, allowing the hybrid vehicle's energy management strategy to select the target engine operating point based on real-time power demands. For power-split hybrid systems, since engine speed and wheel speed can be decoupled through a planetary gear mechanism, the optimized engine operating point set can be used to determine the target engine speed and target engine torque, and the engine can be operated at the target engine operating point through motor speed control.

[0028] This embodiment first determines the emission correlation coefficient between effective fuel consumption rate and multiple effective pollutant emission indicators, and then determines the emission weights in reverse based on the degree of correlation, so that the emission weights no longer rely solely on manual experience. By using effective fuel consumption rate, effective pollutant emission indicators, emission weights, and target power deviation evaluation information together to construct a multi-objective optimization function for engine operating points, it can balance fuel economy and multi-pollutant emission control under the constraint of target power demand. By outputting the set of optimized engine operating points, it can provide directly selectable target engine operating points for the energy management strategy of hybrid vehicles, thereby reducing the online optimization calculation burden and improving the overall fuel consumption and emission performance of the vehicle.

[0029] In one embodiment, engine operating data also includes exhaust flow data. Exhaust flow data can be obtained through an engine bench testing system or calculated based on intake flow, air-fuel ratio, fuel flow, or exhaust system measurement data. The exhaust flow data is used to convert pollutant concentration data into pollutant mass flow rate per unit time, and further combined with engine output power to form an effective pollutant emission index.

[0030] Specifically, the engine output power corresponding to each steady-state operating point is determined based on the engine speed and engine torque. The engine output power can be calculated from the engine speed and engine torque. Therefore, each steady-state operating point corresponds to a specific engine output power, which is used for subsequent calculations of effective fuel consumption rate and effective pollutant emission indicators.

[0031] Then, based on the fuel consumption data and engine output power corresponding to each steady-state operating point, the effective fuel consumption rate corresponding to each steady-state operating point is determined. Fuel consumption data can be instantaneous fuel consumption, fuel mass flow rate, or fuel consumption within a preset sampling time. After converting the fuel consumption data to unit output power, the effective fuel consumption rate in g / kWh can be obtained.

[0032] Furthermore, based on the emission data of various pollutants, exhaust flow rate data, and engine output power corresponding to each steady-state operating point, multiple effective pollutant emission indicators are determined for each steady-state operating point. Pollutant emission data can be in ppm concentrations; the mass emission amount of pollutants can be calculated from the exhaust flow rate data; and then, combined with the engine output power, it can be converted into an effective pollutant emission indicator per unit of output power, such as g / kWh. In this way, different pollutant emission indicators and effective fuel consumption rates can participate in subsequent analyses under a unified evaluation scale.

[0033] Then, based on the effective fuel consumption rate and multiple effective pollutant emission indicators, operating condition index data are obtained to characterize the engine's fuel consumption and emission characteristics. This operating condition index data can be organized according to engine speed and engine torque dimensions, or it can be used to create engine effective fuel consumption rate characteristic diagrams and effective pollutant emission characteristic diagrams, such as... Figure 2 As shown, this operating condition data serves as the data foundation for candidate operating point screening, correlation analysis, and multi-objective optimization.

[0034] This embodiment links engine speed, engine torque, fuel consumption data, pollutant emission data, exhaust flow data, and engine output power, which can convert raw fuel consumption and emission test data into effective fuel consumption rate and effective pollutant emission indicators, making different indicators comparable and providing an accurate data foundation for subsequent candidate operating point screening and fuel consumption-emission synergistic optimization.

[0035] In one embodiment, the emission data for multiple pollutants includes total hydrocarbon emissions, carbon monoxide emissions, and nitrogen oxide emissions. These emissions can be simultaneously collected using an exhaust gas analyzer during engine steady-state testing. Since total hydrocarbons, carbon monoxide, and nitrogen oxides correspond to different combustion and emission formation mechanisms, incorporating all three into the operating point optimization process is beneficial for achieving synergistic control of multiple pollutant emissions.

[0036] Specifically, the effective total hydrocarbon emission index is determined based on total hydrocarbon emission data, exhaust flow data, and the corresponding engine output power. Total hydrocarbon emission data can be total hydrocarbon concentration data, which is then combined with exhaust flow data to convert it into total hydrocarbon mass emissions, and finally combined with engine output power to convert it into the effective total hydrocarbon emission index per unit of output power.

[0037] Effective carbon monoxide emission targets are determined based on carbon monoxide emission data, exhaust flow rate data, and corresponding engine output power. Carbon monoxide emission data can be carbon monoxide concentration data, which is then combined with exhaust flow rate data to convert it into carbon monoxide mass emissions, and finally combined with engine output power to convert it into effective carbon monoxide emission targets.

[0038] Effective nitrogen oxide emission targets are determined based on nitrogen oxide emission data, exhaust flow data, and corresponding engine output power. Nitrogen oxide emission data can be nitrogen oxide concentration data, which is then combined with exhaust flow data to convert it into nitrogen oxide mass emissions, and finally combined with engine output power to convert it into effective nitrogen oxide emission targets.

[0039] Then, based on the effective total hydrocarbon emission index, effective carbon monoxide emission index, and effective nitrogen oxide emission index, multiple effective pollutant emission indices are obtained. These multiple effective pollutant emission indices can be used to generate corresponding emission characteristic maps, such as... Figure 2 The effective carbon monoxide emission characteristics, effective nitrogen oxide emission characteristics, and effective total hydrocarbon emission characteristics are shown in the figure and used for subsequent correlation analysis with effective fuel consumption rate.

[0040] This embodiment converts total hydrocarbons, carbon monoxide, and nitrogen oxides into effective pollutant emission indicators, allowing the emission results of different pollutants to be evaluated based on unit output work. This avoids the problem of incomparability between different operating conditions caused by evaluating only concentration data, and provides a data basis for subsequent identification of the correlation differences between different pollutants and effective fuel consumption rate.

[0041] In one embodiment, a candidate operating point set is determined from multiple steady-state operating points based on the effective fuel consumption rate and preset economic screening criteria. The preset economic screening criteria can be determined based on the target engine's fuel economy zone, engineering calibration requirements, or the energy management needs of hybrid vehicles. In a specific example, the preset economic screening criteria can be set to an effective fuel consumption rate of no more than 250 g / kWh.

[0042] Specifically, the effective fuel consumption rate corresponding to each steady-state operating point is compared with a preset fuel consumption rate threshold to obtain the economic assessment result for each steady-state operating point. The economic assessment result can include two categories: meeting the preset economic screening conditions and not meeting the preset economic screening conditions. When the effective fuel consumption rate of a certain steady-state operating point is not greater than the preset fuel consumption rate threshold, the steady-state operating point can be determined to meet the preset economic screening conditions; when the effective fuel consumption rate of a certain steady-state operating point is greater than the preset fuel consumption rate threshold, the steady-state operating point can be determined to not meet the preset economic screening conditions.

[0043] Then, based on the economic assessment results, steady-state operating points that do not meet the preset economic screening conditions are removed from multiple steady-state operating points, resulting in a set of operating points in the economic zone. This removal process avoids operating points with significantly high fuel consumption from participating in subsequent multi-objective optimization, thus ensuring that the optimized operating points do not deviate from the engine's economic operating range.

[0044] Furthermore, based on the engine speed, engine torque, effective fuel consumption rate, and multiple effective pollutant emission indicators corresponding to each economic zone operating point in the economic zone operating point set, a candidate operating point set is determined. Each candidate operating point in the candidate operating point set can include engine speed, engine torque, engine output power, effective fuel consumption rate, effective total hydrocarbon emission indicator, effective carbon monoxide emission indicator, and effective nitrogen oxide emission indicator. The candidate operating point set can be stored in tabular form or in a two-dimensional grid form composed of engine speed and engine torque.

[0045] This embodiment filters steady-state operating points by pre-setting economic screening conditions, which can limit subsequent correlation analysis and operating point optimization to the engine's economic operating range, avoiding the selection of operating points with significantly higher fuel consumption in order to reduce emissions, thus providing a basis for achieving a comprehensive balance between fuel economy and pollutant emissions.

[0046] In one embodiment, based on the effective fuel consumption rate corresponding to each candidate operating point in the candidate operating point set and multiple effective pollutant emission indicators, the emission correlation coefficient between each effective pollutant emission indicator and the effective fuel consumption rate is determined. Since the effective fuel consumption rate and the numerical ranges of different effective pollutant emission indicators may differ, normalization processing can be performed before calculating the correlation.

[0047] Specifically, the effective fuel consumption rate corresponding to each candidate operating point in the candidate operating point set is normalized to obtain normalized fuel consumption data. Normalization can employ a maximum-minimum normalization method, mapping the effective fuel consumption rate in the candidate operating point set to a preset normalization interval, such as 0 to 1. For any effective fuel consumption rate, a linear mapping can be performed based on the maximum and minimum values ​​of the effective fuel consumption rate in the candidate operating point set to obtain the corresponding normalized fuel consumption data.

[0048] Then, the effective pollutant emission indicators corresponding to each candidate working point in the candidate working point set are normalized to obtain multiple normalized emission indicator data. For the effective total hydrocarbon emission indicator, effective carbon monoxide emission indicator, and effective nitrogen oxide emission indicator, a maximum-minimum normalization method can be used to map different pollutant emission indicators to the same normalization interval. Normalization can eliminate the differences in dimensions and orders of magnitude between different indicators, reducing the impact of indicator dimensions on correlation analysis and subsequent optimization.

[0049] Furthermore, based on normalized fuel consumption data and normalized emission index data, linear correlation parameters between each effective pollutant emission index and the effective fuel consumption rate are calculated, resulting in multiple emission correlation coefficients. The linear correlation parameters can be Pearson correlation parameters. For any effective pollutant emission index, the normalized emission index data corresponding to that effective pollutant emission index in the candidate operating point set can be used as one set of sample data, and the normalized fuel consumption data as another set of sample data. The ratio of the covariance to the standard deviation between the two sets of sample data is calculated to obtain the corresponding emission correlation coefficient. For example... Figure 3 As shown, in a specific example, the emission correlation coefficient between the effective carbon monoxide emission index and the effective fuel consumption rate can be 0.72, the emission correlation coefficient between the effective total hydrocarbon emission index and the effective fuel consumption rate can be 0.15, and the emission correlation coefficient between the effective nitrogen oxide emission index and the effective fuel consumption rate can be 0.06.

[0050] This embodiment normalizes the effective fuel consumption rate and effective pollutant emission indicators before performing linear correlation analysis. This objectively quantifies the strength of the correlation between different pollutant emission indicators and the effective fuel consumption rate, providing a data basis for the subsequent determination of emission weights and reducing the subjectivity brought about by manual weighting based on experience.

[0051] In one embodiment, the emission weights corresponding to each effective pollutant emission index are determined based on the emission correlation coefficient. The emission weights represent the degree of influence of different effective pollutant emission indexes on the multi-objective optimization function at the engine operating point. The determination of emission weights follows the principle that the lower the correlation, the higher the emission weight.

[0052] Specifically, the correlation between each effective pollutant emission indicator and the effective fuel consumption rate is determined based on the absolute value of each emission correlation coefficient. Using the absolute value of the emission correlation coefficient allows for a unified evaluation of the correlation strength under both positive and negative correlation scenarios, avoiding interference from the sign of the correlation coefficient in determining the degree of correlation. A larger absolute value of the emission correlation coefficient indicates a higher degree of correlation between the effective pollutant emission indicator and the effective fuel consumption rate; conversely, a smaller absolute value indicates a lower degree of correlation.

[0053] Then, based on the correlation between multiple effective pollutant emission indicators, a ranking of their correlation is determined. For example, among effective carbon monoxide emissions, effective total hydrocarbon emissions, and effective nitrogen oxide emissions, if the correlation between effective carbon monoxide emissions and effective fuel consumption rate is the highest, followed by effective total hydrocarbon emissions, and then effective nitrogen oxide emissions is the lowest, then the correlation ranking can be expressed as effective carbon monoxide emissions being the highest, effective total hydrocarbon emissions being the lowest, and effective nitrogen oxide emissions being the lowest.

[0054] Furthermore, based on the correlation ranking information and the preset reverse weight allocation rule, the initial emission weights corresponding to multiple effective pollutant emission indicators are determined. The preset reverse weight allocation rule can be set as follows: the higher the correlation, the lower the corresponding initial emission weight; the lower the correlation, the higher the corresponding initial emission weight. In a specific example, there is approximately no linear correlation between the effective nitrogen oxide emission indicator and the effective fuel consumption rate, so a higher initial emission weight can be assigned to the effective nitrogen oxide emission indicator; there is a strong correlation between the effective carbon monoxide emission indicator and the effective fuel consumption rate, so a lower initial emission weight can be assigned to the effective carbon monoxide emission indicator.

[0055] Then, the initial emission weights are normalized to obtain the emission weights corresponding to each effective pollutant emission indicator. The normalized emission weights meet the preset weight range, facilitating the inclusion of multiple effective pollutant emission indicators in the emission evaluation items. Through this process, the emission weights can objectively reflect the differences in the correlation between different pollutant emission indicators and effective fuel consumption rates.

[0056] This embodiment determines the degree of correlation by the absolute value of the emission correlation coefficient, and determines the emission weight according to the correlation ranking information and the preset reverse weight allocation rule. This can strengthen the constraint effect of pollutant emission indicators that are weakly correlated or basically unrelated to the effective fuel consumption rate in multi-objective optimization, avoid the problem that simply reducing fuel consumption cannot effectively constrain the emission of such pollutants, and thus improve the stability and objectivity of the synergistic optimization of fuel economy and multi-pollutant emissions.

[0057] In one embodiment, a multi-objective optimization function for the engine operating point is constructed based on the effective fuel consumption rate, multiple effective pollutant emission indicators, emission weights, and target power deviation evaluation information. This multi-objective optimization function can serve as a fitness function or a comprehensive evaluation function in the subsequent operating point optimization process.

[0058] Specifically, fuel economy evaluation terms are determined based on the effective fuel consumption rate. These terms can be represented by the normalized effective fuel consumption rate or by a weighted effective fuel consumption rate. A smaller fuel economy evaluation term indicates better fuel economy at the candidate operating point.

[0059] Then, emission evaluation items are determined based on multiple effective pollutant emission indicators and their corresponding emission weights. Emission evaluation items can be determined based on multiple normalized emission indicator data and their corresponding emission weights. For example, the effective total hydrocarbon emission indicator, effective carbon monoxide emission indicator, and effective nitrogen oxide emission indicator can be normalized separately and then weighted with their corresponding emission weights to obtain emission evaluation items. For effective pollutant emission indicators that are weakly correlated with or essentially unrelated to effective fuel consumption rate, their influence in the emission evaluation items is enhanced due to their higher corresponding emission weights.

[0060] Furthermore, based on the deviation between the engine output power corresponding to the candidate operating point and the target power requirement, target power deviation evaluation information is determined to constrain the deviation of the candidate operating point from the target power requirement. The target power deviation evaluation information can be the absolute value, the squared value, or the power deviation penalty value corrected by a preset penalty coefficient between the engine output power corresponding to the candidate operating point and the target power requirement. A larger target power deviation evaluation information value indicates a lower degree of matching between the candidate operating point and the target power requirement.

[0061] Then, based on the fuel economy evaluation items, emission evaluation items, and target power deviation evaluation information, a multi-objective optimization function for the engine operating point is constructed. This multi-objective optimization function can be obtained by weighted combination of the fuel economy evaluation items, emission evaluation items, and target power deviation evaluation information. This function allows candidate operating points to achieve a better overall evaluation result while maintaining low fuel consumption, optimal pollutant emissions, and minimal power deviation.

[0062] This embodiment incorporates fuel economy evaluation items, emission evaluation items, and target power deviation evaluation information into the engine operating point multi-objective optimization function. This avoids the deterioration of pollutant emissions caused by pursuing only low fuel consumption, and also avoids the mismatch of target power caused by pursuing only emission improvement. Thus, it achieves comprehensive optimization of fuel consumption and emissions at the engine operating point under the constraint of target power demand.

[0063] In one embodiment, based on a multi-objective optimization function for engine operating points, engine operating points are optimized for multiple target power requirements within a candidate operating point set to obtain the target engine operating point corresponding to each target power requirement. For example... Figure 4As shown, the engine operating point optimization process can be repeatedly executed for multiple target power requirements to obtain an optimized set of engine operating points covering different power requirements.

[0064] Specifically, engine operating point search data is constructed based on the engine speed, engine torque, effective fuel consumption rate, multiple effective pollutant emission indicators, and engine output power corresponding to each candidate operating point in the candidate operating point set. The engine operating point search data can be organized in a gridded format, for example, dividing the engine speed and engine torque dimensions into a preset number of grids to form a two-dimensional operating condition search space. In a specific example, engine power, effective fuel consumption rate, effective total hydrocarbon emissions, effective carbon monoxide emissions, and effective nitrogen oxide emissions can be interpolated or mapped according to the engine operating condition grid to generate 100×100 grid data. Invalid operating points that do not meet the preset economic screening conditions can be marked as invalid values ​​and excluded from subsequent optimization.

[0065] For any target power requirement, based on the engine operating point multi-objective optimization function and a preset global optimization algorithm, an iterative search is performed on candidate operating points in the engine operating point search data to obtain the candidate optimal operating point during the iteration process. The preset global optimization algorithm can be a particle swarm optimization algorithm, or a genetic algorithm, simulated annealing algorithm, or other algorithms capable of global optimization within the search space. When using a particle swarm optimization algorithm, the velocity and position of the particle swarm can be randomly initialized, and the fitness value of the candidate operating point corresponding to the position of each particle can be calculated according to the engine operating point multi-objective optimization function. Subsequently, the individual optimal position and the swarm's global optimal position are updated, and the particles are iteratively updated according to the particle velocity update rules and particle position update rules. The particle swarm size can be set to 50, and the maximum number of iterations can be set to 100; other parameters can also be set according to computational resources and optimization accuracy requirements.

[0066] If the candidate optimal operating point meets the preset convergence conditions, the candidate optimal operating point is determined as the target engine operating point corresponding to the target power requirement. The preset convergence conditions may include reaching the maximum number of iterations, the fitness change being less than a preset change threshold across multiple consecutive iterations, the stability of the group's global optimal position, or the comprehensive evaluation result meeting preset requirements. The target engine operating point includes the target engine speed and target engine torque, and corresponds to performance indicators such as effective fuel consumption rate, effective total hydrocarbon emissions, effective carbon monoxide emissions, and effective nitrogen oxide emissions.

[0067] Furthermore, by traversing multiple target power requirements, the target engine operating point corresponding to each target power requirement is obtained. Multiple target power requirements may include multiple power points within the range of 10kW to 70kW, for example, 13 target power points set at preset power intervals. By performing the above optimization process for each target power requirement, a set of target engine operating points covering different power requirements can be obtained.

[0068] This embodiment constructs engine operating point search data and uses a preset global optimization algorithm for iterative search, which can fully search for engine operating points with better overall performance in the candidate operating point set; by introducing target power deviation evaluation information in the optimization process, it can ensure that the target engine operating point matches the target power requirement; by traversing multiple target power requirements, an optimized engine operating point set applicable to different power requirement scenarios of hybrid vehicles can be formed.

[0069] In one embodiment, the optimized engine operating point set is used by the hybrid vehicle's energy management strategy to select a target engine operating point based on real-time power demand. The optimized engine operating point set can be stored in the vehicle controller as lookup table data, calibration data, or a set of control parameters. Each target engine operating point can correspond to a target power demand and includes the target engine speed and target engine torque.

[0070] Specifically, real-time operating status information of hybrid vehicles is obtained. This real-time operating status information includes real-time power demand and battery state of charge, and may also include vehicle speed, accelerator pedal opening, brake pedal status, engine start / stop status, electric motor operating status, and vehicle drive mode. Real-time power demand can be wheel-side power demand, engine-side power demand, or engine target power demand calculated through energy management strategies.

[0071] Then, based on the state of charge (SBC) of the power battery, it is determined whether the hybrid vehicle is in the battery maintenance phase, thus obtaining the energy management phase judgment result. The battery maintenance phase can be determined based on the relationship between the power battery SBC and a preset SBC threshold. For example, when the power battery SBC is lower than the preset SBC threshold, or when the vehicle control strategy determines that the engine needs to participate in maintaining the battery charge, it can be determined that the hybrid vehicle is in the battery maintenance phase.

[0072] When the energy management phase indicates that the hybrid vehicle is in the battery maintenance phase, a target engine operating point is selected from the set of optimized engine operating points based on the real-time power demand. This target power demand and the real-time power demand satisfy a preset matching condition. The preset matching condition can be either minimizing the difference between the target power demand and the real-time power demand, or ensuring that the difference is less than a preset power error threshold. When multiple target engine operating points satisfy the preset matching condition exist, further selection can be made based on fuel economy evaluation results, emission evaluation results, or vehicle smoothness requirements.

[0073] Then, based on the matched target engine operating point, target engine speed and target engine torque commands are generated to ensure that the hybrid vehicle's engine operates according to these commands. For power-split hybrid systems, a planetary gear mechanism can be used to decouple engine speed from wheel speed, and the actual engine speed can track the target engine speed through generator speed control, while the engine controller outputs the target torque. Thus, the engine can operate at the pre-optimized target engine operating point throughout the vehicle's operation.

[0074] This embodiment applies the optimized engine operating point set to the energy management strategy of hybrid vehicles, which avoids online global optimization of the engine operating space in each control cycle during vehicle operation, reducing the computational burden on the controller; by selecting the target engine operating point from the optimized engine operating point set according to real-time power demand, the adaptability of the engine operating point to the vehicle power demand can be improved; by controlling the engine operation according to the target engine speed and target engine torque during the battery maintenance phase, fuel consumption and pollutant emission control can be taken into account in the whole vehicle cycle.

[0075] In one embodiment, a multi-objective optimization method for engine operating point based on inverse adjustment of correlation coefficient is provided, and its specific implementation steps are as follows: Step 1: Engine bench testing and data acquisition: A target engine (in this embodiment, a 2.0L naturally aspirated hybrid gasoline engine) was selected, and steady-state performance tests were conducted on an engine test bench across the entire operating range. The test speed range covered the engine's commonly used operating range (e.g., 1000~5600 r / min), and the test load range covered 0~100% throttle opening. At each steady-state operating point, engine speed, torque, instantaneous fuel consumption, and the concentrations of various pollutants in the exhaust gas, including total hydrocarbons (THC), carbon monoxide (CO), and nitrogen oxides (NOx), were simultaneously collected. Fuel consumption was measured using a transient fuel flow meter, and pollutant concentrations were collected using an exhaust gas analyzer. The collected raw data were used for subsequent analysis.

[0076] Step 2: Data Preprocessing and Dimensional Standardization The raw data collected in Step 1 is preprocessed. First, the concentration data of each pollutant (in ppm) are converted into effective pollutant emissions in g / kWh according to the engine's exhaust flow rate and power output, namely, effective total hydrocarbons (BSTHC), effective carbon monoxide (BSCO), and effective nitrogen oxides (BSNOx), so that they have the same dimension as the effective fuel consumption rate (BSFC, in g / kWh). Second, an economic screening threshold is set (BSFC ≤ 250 g / kWh in this embodiment), and the non-economic operating points of the engine are eliminated, retaining only the operating points with relatively high fuel economy as the candidate set of working points for subsequent optimization. The resulting engine BSFC characteristic MAP and effective pollutant emission characteristic MAP are as follows: Figure 3 As shown.

[0077] Then, Max-Min normalization was performed on the screened BSFC and emission data of each effective pollutant, linearly mapping the data range of each indicator to the [0, 1] interval to eliminate the influence of magnitude differences between different indicators on subsequent analysis. Max-Min has advantages such as eliminating dimensional differences, balancing magnitude effects, insensitivity to local outliers, physical intuitiveness, and interpretability. Its calculation formula is as follows:

[0078] In the formula, , , as well as These represent the original value, maximum value, minimum value, and normalized value in the dataset, respectively.

[0079] Step 3: Correlation analysis between fuel consumption and emissions: Using normalized data, Pearson correlation coefficients were calculated between emissions of each pollutant (HC, CO, NOx) and fuel consumption rate (BSFC). Correlation analysis is based on the Pearson correlation coefficient principle, which quantifies the relationship between two variables using the correlation coefficient r. The correlation coefficient r ranges from -1 to 1, and its magnitude directly reflects the closeness of the relationship between the two variables. Its mathematical formula is as follows:

[0080] In the formula, x and y represent any two variables for correlation analysis; , , Let x and y represent the covariance of the x and y samples, the standard deviation of the x sample, and the standard deviation of the y sample, respectively. and Representing variables respectively x and y In the i Specific observations on each sample and Let r and r represent the sample means (i.e. averages) of variables x and y, respectively. The correlation coefficient has the following characteristics: when r = 1, there is a perfect positive correlation; when r = -1, there is a perfect negative correlation; when r = 0, there is no linear correlation.

[0081] Based on the calculation results, the linear correlation strength between each pollutant and BSFC was obtained: CO showed a strong positive correlation with BSFC (r=0.72), HC showed a weak positive correlation with BSFC (r=0.15), while NOx showed almost no linear correlation with BSFC (r=0.06). This correlation coefficient will serve as the quantitative basis for subsequent weight inverse adjustment.

[0082] Step 4: Construct a multi-objective optimization function based on inverse adjustment of correlation coefficient: Based on the correlation coefficients between each pollutant and BSFC obtained in step three, a multi-objective optimization function is designed. The core of this embodiment is that pollutants with weaker correlation to BSFC should be assigned higher weights in the multi-objective optimization function, i.e., "reverse adjustment." The optimization objective function consists of three objective terms: the first term is the BSFC term, the second term is the joint emission term related to BSFC (including HC and CO), and the third term is the independent emission term unrelated to BSFC (i.e., NOx). The objective function is defined as follows: ; In the formula, , , , These represent the normalized BSFC, HC, CO, and NOx data, respectively. , , These represent the weights of the three objectives (fuel economy, BSFC-related joint emissions, and BSFC-independent independent emissions). , These represent the individual weights of HC and CO within the second term (BSFC related term), and their calculation formulas are shown below:

[0083] In the formula, , These are the correlation coefficients of HC, CO, and BSFC, respectively. The weighting formula adjusts the impact on the optimization direction based on the correlation between objectives, which is a reverse adjustment mechanism. Emissions with lower correlation to BSFC indicate higher objective independence and will be assigned a higher weight in the objective function. In this way, the weight of strongly correlated BSFC terms can be weakened while the influence of weakly correlated BSFC terms in the objective function is strengthened, thereby achieving an overall balance between fuel economy and emission control.

[0084] Step 5: Optimizing the operating point based on the particle swarm optimization algorithm: Within the candidate operating point set (BSFC ≤ 250 g / kWh) determined in step two, the multi-objective optimization function constructed in step four is used as the fitness function. Particle Swarm Optimization (PSO) algorithm is employed to traverse and optimize the engine's optimal operating point under different target power points. The PSO algorithm searches for the optimal engine operating point under different target power levels, i.e., the speed-torque operating point. The engine's power, BSFC, and effective pollutant emissions (BSHC, BSCO, BSNOx) MAPs are gridded to generate 100×100 grid data. The PSO algorithm is used to search within the engine operating point grid, employing four different weight configuration schemes and incorporating power deviation as a penalty term to comprehensively balance fuel economy (BSFC), emission indicators (THC, CO, NOx), and power deviation. Row index "i" corresponds to the speed dimension, and column index "j" corresponds to the torque dimension. Invalid operating points (BSFC > 250 g / kWh) are marked as NaN, and performance indicators are normalized to the [0, 1] range using Max-Min. The preprocessed engine data is organized in the following matrix structure: ; The PSO population size is set to 50, the maximum number of iterations is 100, and the particle position and velocity update formulas are as follows: Particle position update: ; Particle velocity update: ; In the formula, and These represent the particle's current velocity and the velocity of its next iteration, respectively. and These represent the particle's current and next iteration positions, respectively; ω is the inertia weight. c 1 represents the individual moderating weight. c 2 represents the social adjustment weight. r 1 and r 2 is a random number in the range [0, 1]. p_best i and g_besti These represent the current best position of an individual and the best position in the history of the population, respectively.

[0085] The fitness function of the Particle Swarm Optimization (PSO) algorithm is constructed based on the multi-objective optimization function built in step four. By introducing a power point deviation penalty term, the deviation between the search candidate point and the target power point is controlled. The fitness function is shown in the following equation: +

[0086] In the formula, Represents grid points ( i, j The corresponding engine operating point adaptability value; the smaller the value, the better the overall performance of that operating point in terms of fuel consumption, emissions, and power matching. , , These represent the weights of the three objectives mentioned in the objective function definition in "Step Four" (fuel economy, BSFC-related joint emissions, and BSFC-independent independent emissions). , These represent the weights assigned to HC and CO separately within the second item (BSFC related item); , , , Representing grid points ( i, j The corresponding normalized BSFC, HC, CO, and NOx data; PowerPenalty The power point offset penalty term is calculated using grid points ( i, j The corresponding engine output power The target power point specified in the current optimization task The absolute value of the deviation is multiplied by a manually set penalty factor coefficient (0.1 in this example). The calculation formula is as follows:

[0087] For the 13 target power points (10~70 kW) of the engine, the engine operating point optimization based on the PSO algorithm is carried out sequentially for schemes one to four. The specific process is as follows: Randomly initialize the velocity and position of the particle swarm; The fitness value of each candidate operating point is calculated based on the multi-objective optimization function of the engine operating point. The multi-objective optimization function of the engine operating point includes fuel economy evaluation items, emission evaluation items and target power deviation evaluation information, so as to evaluate the comprehensive performance of the engine operating point corresponding to the particle's location. Update individual optimal solutionp_best and the global optimal solution of the population g_best ; Update the particle's velocity and position; The above process is repeated iteratively until the convergence condition is met or the maximum number of iterations is reached. Iterate through the four schemes, output and record the location index of each optimal target power point in turn; The engine speed, torque value, and various performance indicators at the operating point can be found by looking up the corresponding position in the table using the index, including effective fuel consumption (BSFC) and effective pollutant emissions (BSHC, BSCO, and BSNOx).

[0088] Step Six: Output the optimized engine operating point set: By iterating through all target power points and repeating the optimization process in step five, a set of engine operating points with optimal overall performance under different power requirements is finally obtained (each operating point includes speed and torque information). The engine operating points in this set are all located within the engine's economic operating range, and with almost no increase in fuel consumption, the original HC emissions in the low-power range of the engine are reduced, and the original NOx emissions (full power range) are significantly improved, achieving a better balance between fuel economy and pollutant emissions.

[0089] Step 7: Optimize the application of the operating point set in hybrid power energy management strategy: The optimized engine operating point set output from step six is ​​used as candidate operating points and integrated into the rule-based energy management strategy of the hybrid vehicle. Specifically, when the vehicle is in a low-charge maintenance phase (e.g., when the battery SOC is below a preset threshold), the rule-based energy management strategy based on the optimized operating points selects the engine speed and torque corresponding to the power point closest to the demand from the optimized operating point set as the target command for the engine, according to the real-time wheel power demand. For the power-split hybrid system, the decoupling capability of its planetary gear mechanism between engine speed and wheel speed is utilized, and the speed control of the generator (MG1) ensures that the engine actually operates at the target operating point, thereby achieving comprehensive optimization of fuel economy and pollutant emissions during the cycle.

[0090] This embodiment optimizes engine operating points, achieving better overall performance in terms of fuel consumption and emissions levels at various operating conditions. Specifically, the comparison of the optimization results of the multi-objective optimization method designed in this embodiment with the "single optimization method with the objective of minimizing fuel consumption" is as follows: In terms of effective fuel consumption (BSFC), there were only slight increases of 1.8 g / kWh and 2.3 g / kWh at the 15 kW and 65 kW power points, respectively, with increases of less than 1%. At other power points, there were significant differences between the two. Regarding effective total hydrocarbon (BSTHC) emissions, this embodiment shows improvements of 6.3% and 7.0% at the 10 kW and 15 kW power points, respectively. In terms of effective carbon monoxide (BSCO) emissions, the two have their own advantages and disadvantages at different power points; In terms of effective nitrogen oxide (BSNOx) emissions, this embodiment outperforms the "single optimization method with the goal of minimizing fuel consumption" at almost all power points, with the most significant reduction at the 65 kW power point, with an improvement of up to 19.7%.

[0091] When applied to energy management strategies, it improves the overall performance of cumulative fuel consumption and emission control within the vehicle cycle. The optimization method of this embodiment and the "single optimization method with the goal of minimizing fuel consumption" are integrated into the whole vehicle CLTC-C (China Light Commercial Vehicle Test Cycle) cycle simulation after the rule-based strategy. The fuel consumption and the cumulative emissions of various pollutants are compared. The results show that: Regarding cumulative fuel consumption, with the SOC consistent at the beginning and end of the cycle, the cumulative fuel consumption using the strategies of "this embodiment" and "single target of minimum fuel consumption" were 882.2 g and 883.4 g, respectively, with negligible differences. In terms of cumulative HC emissions, this embodiment shows an in-cycle improvement of 0.44%. Regarding cumulative CO emissions, the degradation within the cycle in this embodiment is less than 0.14%, which is negligible. In terms of cumulative NOx emissions, this embodiment shows an in-cycle improvement of 1.80%.

[0092] This embodiment uses Pearson correlation analysis to scientifically quantify the strength of the correlation between the generation of three types of pollutants (HC, CO, and NOx) and BSFCs, and designs a multi-objective optimization function with inversely adjusted weights based on this. This method overcomes the subjectivity of weight setting in traditional multi-objective optimization, making the optimization results objective and reproducible. The optimized engine operating point in this embodiment, compared with the optimization method that "solely aims at economy", achieves a moderate improvement in HC emissions and a significant improvement in NOx emissions with only a slight increase in fuel consumption at certain operating points. The optimal speed-torque operating points at different power levels determined in this embodiment can be decoupled from the engine speed and wheel speed by utilizing the planetary gear mechanism of the power split hybrid power system, enabling the engine's actual operating point to accurately track the optimization target, which has good engineering applicability. Applying the optimized operating point of this embodiment to a regular hybrid power energy management strategy can effectively improve the overall fuel consumption and emission performance of the vehicle, providing a basis for candidate engine operating points that balance fuel consumption and emissions for hybrid power systems.

[0093] In one embodiment, such as Figures 2 to 4 As shown, a multi-objective optimization system for engine operating points suitable for hybrid vehicles is provided. This system may include a data acquisition module, an index determination module, a correlation determination module, a weight determination module, a function construction module, an operating point optimization module, and an operating point set generation module. Each module can be implemented by a vehicle controller, an engine controller, a calibration calculation platform, an offline optimization platform, or a combination thereof, or by a processor executing program instructions from memory.

[0094] The data acquisition module is used to acquire engine operating condition data of the target engine at multiple steady-state operating points. Engine operating condition data includes engine speed, engine torque, fuel consumption data, and various pollutant emission data. In some embodiments, the engine operating condition data may also include exhaust flow rate data. The data acquisition module can communicate with an engine bench testing system, fuel flow meter, exhaust gas analyzer, and data acquisition equipment to acquire steady-state operating condition test data.

[0095] The indicator determination module is used to determine the effective fuel consumption rate and multiple effective pollutant emission indicators corresponding to each steady-state operating point based on engine operating condition data. It then determines a set of candidate operating points from these steady-state operating points based on the effective fuel consumption rate and preset economic screening criteria. This module can convert fuel consumption data into an effective fuel consumption rate and combine multiple pollutant emission data with exhaust flow data and engine output power to convert them into multiple effective pollutant emission indicators. Furthermore, the module can filter economic zone operating points based on preset fuel consumption rate thresholds to form a set of candidate operating points.

[0096] The correlation determination module is used to determine the emission correlation coefficients between each effective pollutant emission index and the effective fuel consumption rate, based on the effective fuel consumption rate and multiple effective pollutant emission indices corresponding to each candidate operating point in the candidate operating point set. The correlation determination module can normalize the effective fuel consumption rate and multiple effective pollutant emission indices, and calculate linear correlation parameters based on the normalized data to obtain multiple emission correlation coefficients.

[0097] The weight determination module is used to determine the emission weights corresponding to each effective pollutant emission indicator based on the emission correlation coefficient. The lower the correlation between the effective pollutant emission indicator and the effective fuel consumption rate, the higher the emission weight corresponding to the effective pollutant emission indicator. The weight determination module can determine the correlation degree based on the absolute value of the emission correlation coefficient, determine the initial emission weights based on the correlation degree ranking information and the preset reverse weight allocation rule, and normalize the initial emission weights to obtain the final emission weights.

[0098] The function construction module is used to construct a multi-objective optimization function for the engine operating point based on effective fuel consumption rate, multiple effective pollutant emission indicators, emission weights, and target power deviation evaluation information. The module can generate fuel economy evaluation items, emission evaluation items, and target power deviation evaluation items, and combine these evaluation items into a fitness function or a comprehensive evaluation function.

[0099] The operating point optimization module is used to optimize the engine operating point for multiple target power requirements based on a multi-objective optimization function of the engine operating point. It optimizes the engine operating point for each target power requirement within a candidate operating point set. The module can construct engine operating point search data and perform iterative searches based on preset global optimization algorithms such as particle swarm optimization, thereby outputting the target engine speed and target engine torque.

[0100] The operating point set generation module generates an optimized engine operating point set based on the target engine operating points corresponding to multiple target power demands. This optimized operating point set is used by the hybrid vehicle's energy management strategy to select the target engine operating point based on real-time power requirements. The operating point set generation module can associate and store multiple target power demands with corresponding target engine speeds, target engine torques, and related performance indicators for subsequent vehicle control calls.

[0101] This embodiment achieves a complete processing flow from engine bench data to an optimized engine operating point set through data transfer between the data acquisition module, index determination module, correlation determination module, weight determination module, function construction module, operating point optimization module, and operating point set generation module. The correlation determination and weight determination modules enable objective determination of emission weights based on the fuel consumption-emission correlation. The function construction and operating point optimization modules allow for obtaining a target engine operating point that balances fuel economy and multi-pollutant emissions under target power demand constraints. The operating point set generation module provides a directly callable foundation of optimized operating points for hybrid vehicle energy management strategies.

[0102] 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.

Claims

1. A multi-objective optimization method for engine operating point in hybrid vehicles, characterized in that, include: The engine operating condition data of the target engine at multiple steady-state operating points are obtained. The engine operating condition data includes engine speed, engine torque, fuel consumption data and various pollutant emission data. Based on the engine operating data, the effective fuel consumption rate and multiple effective pollutant emission indicators corresponding to each steady-state operating point are determined, and a candidate operating point set is determined from multiple steady-state operating points based on the effective fuel consumption rate and preset economic screening conditions. Based on the effective fuel consumption rate corresponding to each candidate operating point in the candidate operating point set and multiple effective pollutant emission indicators, determine the emission correlation coefficient between each effective pollutant emission indicator and the effective fuel consumption rate; The emission weights corresponding to each effective pollutant emission index are determined based on the emission correlation coefficients; wherein, the lower the correlation between the effective pollutant emission index and the effective fuel consumption rate, the higher the emission weight corresponding to the effective pollutant emission index. Based on the effective fuel consumption rate, multiple effective pollutant emission indicators, emission weights, and target power deviation evaluation information, a multi-objective optimization function for engine operating point is constructed. Based on the multi-objective optimization function of the engine operating point, the engine operating point is optimized for multiple target power requirements in the candidate operating point set to obtain the target engine operating point corresponding to each target power requirement; wherein, the target engine operating point includes the target engine speed and the target engine torque. Based on the target engine operating points corresponding to the multiple target power requirements, an optimized engine operating point set is generated. This optimized engine operating point set is used by the hybrid vehicle energy management strategy to select the target engine operating point based on real-time power requirements.

2. The multi-objective optimization method for engine operating point in hybrid vehicles according to claim 1, characterized in that, The engine operating data also includes exhaust flow data; The step of determining the effective fuel consumption rate and multiple effective pollutant emission indicators corresponding to each steady-state operating point based on the engine operating data includes: The engine output power corresponding to each steady-state operating point is determined based on the engine speed and engine torque corresponding to each steady-state operating point. Based on the fuel consumption data and engine output power corresponding to each steady-state operating point, determine the effective fuel consumption rate corresponding to each steady-state operating point; Based on the emission data of various pollutants, exhaust flow data and engine output power corresponding to each steady-state operating point, determine multiple effective pollutant emission indicators corresponding to each steady-state operating point. Based on the effective fuel consumption rate and multiple effective pollutant emission indicators, operating condition index data for characterizing engine fuel consumption and emission characteristics are obtained.

3. The multi-objective optimization method for engine operating point in hybrid vehicles according to claim 2, characterized in that, The emission data for the various pollutants include total hydrocarbon emission data, carbon monoxide emission data, and nitrogen oxide emission data; The step of determining multiple effective pollutant emission indicators corresponding to each steady-state operating point based on various pollutant emission data, exhaust flow data, and engine output power includes: The effective total hydrocarbon emission index is determined based on the total hydrocarbon emission data, the exhaust flow data, and the corresponding engine output power. The effective carbon monoxide emission index is determined based on the carbon monoxide emission data, the exhaust flow data, and the corresponding engine output power. The effective nitrogen oxide emission index is determined based on the nitrogen oxide emission data, the exhaust flow data, and the corresponding engine output power. Based on the effective total hydrocarbon emission index, the effective carbon monoxide emission index, and the effective nitrogen oxide emission index, multiple effective pollutant emission indices are obtained.

4. The multi-objective optimization method for engine operating point in hybrid vehicles according to claim 2, characterized in that, The step of determining a candidate operating point set from multiple steady-state operating points based on the effective fuel consumption rate and preset economic screening conditions includes: The effective fuel consumption rate corresponding to each steady-state operating point is compared with the preset fuel consumption rate threshold to obtain the economic judgment result corresponding to each steady-state operating point. Based on the economic assessment results, steady-state operating points that do not meet the preset economic screening conditions are removed from the multiple steady-state operating points to obtain the set of operating points in the economic zone. The candidate operating point set is determined based on the engine speed, engine torque, effective fuel consumption rate, and multiple effective pollutant emission indicators corresponding to each operating point in the economic zone operating point set.

5. The multi-objective optimization method for engine operating point in hybrid vehicles according to claim 1, characterized in that, The step of determining the emission correlation coefficient between each effective pollutant emission index and the effective fuel consumption rate based on the effective fuel consumption rate corresponding to each candidate operating point in the candidate operating point set and multiple effective pollutant emission indices includes: The effective fuel consumption rate corresponding to each candidate operating point in the candidate operating point set is normalized to obtain normalized fuel consumption data. The multiple effective pollutant emission indicators corresponding to each candidate working point in the candidate working point set are normalized to obtain multiple normalized emission indicator data. Based on the normalized fuel consumption data and the normalized emission index data, the linear correlation parameters between each effective pollutant emission index and the effective fuel consumption rate are calculated to obtain multiple emission correlation coefficients.

6. The multi-objective optimization method for engine operating point in hybrid vehicles according to claim 5, characterized in that, The step of determining the emission weight corresponding to each effective pollutant emission indicator based on the emission correlation coefficient includes: The degree of correlation between each effective pollutant emission index and the effective fuel consumption rate is determined based on the absolute value of each emission correlation coefficient. Based on the correlation degree of the multiple effective pollutant emission indicators, determine the correlation degree ranking information of the multiple effective pollutant emission indicators; Based on the correlation ranking information and the preset reverse weight allocation rule, the initial emission weights corresponding to multiple effective pollutant emission indicators are determined; The initial emission weights are normalized to obtain the emission weights corresponding to each effective pollutant emission index; wherein, the emission weights corresponding to effective pollutant emission indices with lower correlation are higher.

7. The multi-objective optimization method for engine operating point in hybrid vehicles according to claim 1, characterized in that, The step of constructing a multi-objective optimization function for the engine operating point based on the effective fuel consumption rate, multiple effective pollutant emission indicators, emission weights, and target power deviation evaluation information includes: Based on the effective fuel consumption rate, determine the fuel economy evaluation items; Based on the multiple effective pollutant emission indicators and their corresponding emission weights, emission evaluation items are determined; Based on the deviation between the engine output power corresponding to the candidate operating point and the target power requirement, target power deviation evaluation information is determined to constrain the deviation of the candidate operating point from the target power requirement. Based on the fuel economy evaluation item, the emission evaluation item, and the target power deviation evaluation information, a multi-objective optimization function for the engine operating point is constructed.

8. The multi-objective optimization method for engine operating point in hybrid vehicles according to claim 1, characterized in that, The multi-objective optimization function based on the engine operating point optimizes the engine operating point for multiple target power requirements in the candidate operating point set to obtain the target engine operating point corresponding to each target power requirement, including: Based on the engine speed, engine torque, effective fuel consumption rate, multiple effective pollutant emission indicators, and engine output power corresponding to each candidate operating point in the candidate operating point set, engine operating point search data is constructed. For any target power requirement, based on the engine operating point multi-objective optimization function and the preset global optimization algorithm, the candidate operating points in the engine operating point search data are iteratively searched to obtain the candidate optimal operating point in the iterative process. If the candidate optimal operating point satisfies the preset convergence condition, the candidate optimal operating point is determined as the target engine operating point corresponding to the target power requirement; By iterating through multiple target power requirements, the target engine operating point corresponding to each target power requirement is obtained.

9. The multi-objective optimization method for engine operating point in hybrid vehicles according to claim 8, characterized in that, The engine optimization operating point set is used by the hybrid vehicle energy management strategy to select the target engine operating point based on real-time power demand, including: Obtain real-time operating status information of hybrid vehicles, including real-time power demand and power battery state of charge; Based on the state of charge of the power battery, determine whether the hybrid vehicle is in the power maintenance phase, and obtain the energy management phase judgment result; When the energy management phase judgment result indicates that the hybrid vehicle is in the battery maintenance phase, a target engine operating point whose target power demand meets the preset matching condition with the real-time power demand is selected from the set of optimized engine operating points based on the real-time power demand. Based on the matched target engine operating point, a target engine speed command and a target engine torque command are generated so that the engine of the hybrid vehicle operates in accordance with the target engine speed command and the target engine torque command.

10. A multi-objective optimization system for engine operating point in hybrid vehicles, characterized in that, include: The data acquisition module is used to acquire engine operating condition data of the target engine at multiple steady-state operating points. The engine operating condition data includes engine speed, engine torque, fuel consumption data, and various pollutant emission data. The indicator determination module is used to determine the effective fuel consumption rate and multiple effective pollutant emission indicators corresponding to each of the engine operating condition points based on the engine operating condition data, and to determine a set of candidate operating points from the multiple steady-state operating points based on the effective fuel consumption rate and preset economic screening conditions. The correlation determination module is used to determine the emission correlation coefficient between each effective pollutant emission index and the effective fuel consumption rate based on the effective fuel consumption rate corresponding to each candidate operating point in the candidate operating point set and multiple effective pollutant emission indicators. The weight determination module is used to determine the emission weight corresponding to each effective pollutant emission index based on the emission correlation coefficient; wherein, the lower the correlation between the effective pollutant emission index and the effective fuel consumption rate, the higher the emission weight corresponding to the effective pollutant emission index. The function construction module is used to construct a multi-objective optimization function for the engine operating point based on the effective fuel consumption rate, multiple effective pollutant emission indicators, the emission weights, and target power deviation evaluation information. The operating point optimization module is used to optimize the engine operating point for multiple target power requirements in the candidate operating point set based on the engine operating point multi-objective optimization function, and obtain the target engine operating point corresponding to each target power requirement; wherein, the target engine operating point includes the target engine speed and the target engine torque; The operating point set generation module is used to generate an optimized engine operating point set based on the target engine operating points corresponding to the multiple target power requirements. The optimized engine operating point set is used for the hybrid vehicle energy management strategy to select the target engine operating point according to the real-time power requirements.

Images