Range extender mode switching control method considering energy consumption emission and battery life

Abstract

The invention is suitable for the technical field of energy management of extended-range electric vehicles, and provides a range extender mode switching control method considering energy consumption emission and battery life, the method comprises the following steps: collecting vehicle driving data in real time, using a pulse noise filter and a smooth curve filter to carry out noise reduction pretreatment on a vehicle speed sequence, and obtaining a range extender mode switching control result; and based on a fuzzy algorithm and a variable scale window algorithm, a driving style evaluation coefficient and a road condition passing degree evaluation coefficient are identified on line. And dynamically determining a mode switching threshold value in combination with the comprehensive cost function and the pre-driving mileage to realize intelligent switching of a top energy mode. The working mode of the APU system is dynamically adjusted, high-power output is preferentially performed under aggressive driving and congested road conditions, and efficient operation is optimized under stable driving and smooth road conditions. Through a multi-objective optimization strategy, the energy consumption of the whole vehicle is obviously reduced by 8%-12%, emission is reduced, the service life of a battery is prolonged by 20% or above, and the method is suitable for range-extended electric vehicle energy management under complex urban road conditions.

Description

Technical Field

[0001] The present invention belongs to the technical field of energy management of extended-range electric vehicles, and in particular relates to a range extender mode switching control method that takes energy consumption, emissions and battery life into consideration. Background Art

[0002] Extended-range electric vehicles (EREVs) have become an important technical path to alleviate range anxiety because they combine the cleanliness of pure electric drive with the flexibility of range extender energy replenishment. Its core advantage is that through the coordinated work of the engine (range extender) and the power battery, the engine can be freed from the constraints of complex working conditions, focusing on high-efficiency range operation, reducing fuel consumption and emissions.

[0003] Existing energy management strategies for extended-range electric vehicles are mostly based on fixed rules or offline optimization algorithms, which do not fully incorporate real-time driving style and road condition information, leading to the following problems: Energy consumption and emissions are high. Traditional strategies cannot dynamically adjust the APU (range extender) working mode under complex road conditions, resulting in poor fuel economy. Battery life decays quickly, and frequent charging and discharging and drastic power fluctuations accelerate battery capacity decay. Mode switching response is delayed, and the parameter recognition algorithm of a single scale window is difficult to adapt to different driving conditions (such as congested and unobstructed sections), resulting in inaccurate mode switching. Summary of the invention

[0004] The object of the present invention is to provide a range extender mode switching control method taking energy consumption emissions and battery life into consideration, aiming to solve the technical problems existing in the prior art identified in the background technology.

[0005] The present invention is implemented as follows: a range extender mode switching control method considering energy consumption emissions and battery life, characterized in that the method includes: Collect vehicle driving data in real time, perform noise reduction preprocessing on the vehicle speed sequence data through impulse noise filter and smooth curve filter, divide the working condition segments and extract characteristic parameters; Based on the vehicle speed sequence data and the preset mileage, the driving style evaluation coefficient is identified online based on the fuzzy algorithm and the variable scale window algorithm. EC ds and road condition passability evaluation coefficient EC rt , and optimize key threshold parameters; Defining the comprehensive cost function F c and switch parameters k cr , combined with real-time battery SoC and estimated driving range M preset , switch the top energy modes CD-EV+CS-Blend and CD-Blend; Dynamically switch the APU system working mode according to driving style and road conditions.

[0006] The beneficial effects of the present invention are: The present invention selects a driving style evaluation index to evaluate driving style, selects road congestion to evaluate real-time road conditions, and realizes the recognition of driving style and road condition characteristics based on a fuzzy algorithm. The experimental results verify that the research adopts a driving style and road condition characteristic recognition algorithm based on a fuzzy algorithm. The pattern recognizer can better complete the online recognition function through "general fuzzification" on information type segmentation.

[0007] The present invention designs a parameter recognition algorithm based on a variable scale window to obtain the vehicle speed sequence information of the working condition fragment in real time, and applies it to the "mode switching" of the real-time energy management strategy. After the key parameters in the algorithm are optimized, a high-precision recognition result can be obtained. The energy management strategy formulated responds accurately to the mode switching made by the changes in the driving characteristics of the vehicle, and realizes good recognition and online application of travel characteristic information. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 is a flow chart of a range extender mode switching control method considering energy consumption, emission and battery life; Figure 2 It is the speed sequence diagram of the driving route; Figure 3 This is the result diagram of the impact standard deviation identification under the driving cycle condition; Figure 4 is the membership function and output surface of the fuzzy controller Figure 1 ; Figure 5 is the membership function and output surface of the fuzzy controller Figure 2 ; Figure 6 The average value of the optimization variable during the iteration process Figure 1 ; Figure 7 The average value of the optimization variable during the iteration process Figure 2 ; Figure 8 The average value of the optimization variable during the iteration process Figure 3 ; Fig. 9 The average value of the optimization variable during the iteration process Figure 4 ; Fig.10 This is the result diagram of the impact standard deviation identification under the driving cycle condition; Fig.11 Identify the result graph of driving style assessment coefficients; Fig.12Implement a flow chart for an intelligent “mode switching” energy management strategy based on travel characteristics; Fig.13 The statistical results of comprehensive driving cost and mileage under different strategies Figure 1 ; Fig.14 The statistical results of comprehensive driving cost and mileage under different strategies Figure 2 ; Fig.15 The statistical results of comprehensive driving cost and mileage under different strategies Figure 3 ; Fig.16 The statistical results of comprehensive driving cost and mileage under different strategies Figure 4 ; Fig.17 For CD-Blend strategy and CD-EV+CS-Blend strategy F c Function intersection distribution result diagram; Fig.18 This is a schematic diagram of the APU system working mode switching rules under the CD-EV+CS-Blend strategy; Fig.19 This is a schematic diagram of the APU system working mode switching rules under the CD-Blend strategy; Fig. 20 Schematic diagram of the APU system working mode switching rules. DETAILED DESCRIPTION

[0009] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0010] like Figure 1 As shown, a range extender mode switching control method considering energy consumption, emission and battery life includes: Step 1: Based on the fuzzy algorithm, the speed sequence data of the working condition fragment in a period of time before the real-time driving process is occurring is obtained and the noise reduction preprocessing is performed; The average traffic flow statistics method is used to collect driving cycle data. The driver drives the car following the current traffic flow on the planned test route. The real vehicle data is collected using USB-CAN-OBD equipment with a sampling time of 1s to obtain various values ​​of vehicle operation in real time, including vehicle speed, throttle opening, engine speed, etc. The choice of the test route is directly related to the authenticity of the constructed driving conditions. Considering factors such as the test cycle, test cost, and the ability of the test results to reflect the real conditions of various types of roads, the collected driving condition basic data is targeted at specific cities, and the road areas are divided into the prosperous area of ​​the city center, the residential area, and the ring road.

[0011] The invalid speed sequence when the vehicle is waiting for a traffic light and the APU system engine is shut down is removed from the data. The preprocessed data is divided into operating condition segments, and characteristic parameters that can reflect the driving style and driving condition characteristics are extracted from the operating condition segments. An impulse noise filter is used to remove the singular points of the operating condition curve, and then a high-frequency noise filter is used to smooth the operating condition curve.

[0012] The impulse noise filter is defined as follows: ; In the formula, represent The speed of the car at the moment, Represents the driving speed after filtering; Smooth Curve Filter: ; represent The driving speed after filtering at each moment, F Represents the vehicle speed weight.

[0013] like Figure 2 As shown in the figure, the speed sequence data of four driving cycle conditions, CCDC, CC-CR, CU-OR, and CS-SR, are selected as the basic data, and the condition segments are divided based on the micro-stroke condition sample division method. Each condition sample obtained includes the complete process from vehicle start acceleration, deceleration and braking to parking. The four driving cycle data are divided into about 318 condition segments for the subsequent recognition study of driving style and road condition information characteristics.

[0014] Step 2: Identify the driving style evaluation coefficient online based on the obtained speed sequence data and the preset mileage EC ds and road condition passability evaluation coefficient EC rt , and based on the parameter recognition algorithm of the variable scale recognition window, the key threshold parameters are optimized to achieve the accuracy and effectiveness of the recognition results.

[0015] The driving styles of drivers are usually divided into three categories: aggressive, normal and calm. Aggressive drivers step on the accelerator and brake pedals with large amplitudes and high speeds, and the impact changes dramatically. Normal drivers use the accelerator and brake pedals relatively reasonably, and the impact changes moderately. Calm drivers step on the accelerator and brake pedals with small amplitudes and low speeds, and the impact changes slightly. Considering the recognition accuracy and recognition speed, the degree of vehicle acceleration change-impact (how violently the driver steps on the accelerator or brake pedal) is selected to reflect the driving style. The maximum impact cannot represent the general situation of the entire road section. The average impact may be zero when the fluctuation is large during a period of time, and it cannot reflect the drastic degree of impact change during this period. The driver style recognition coefficient is selected. A c R , the evaluation indicators are defined as follows: ; In the formula, SD A c R Represents the standard deviation of the impact degree within a working condition identification cycle, A c R Represents the average absolute value of the typical working condition impact degree of the working condition category of ordinary drivers in a working condition identification cycle, T R represents the length of the recognition cycle, A c R ( t R )represent t R The impact degree identification coefficient at the moment, SD A c R It can reflect the change of impact degree (discrete degree) within a certain period of time. a c R represents the acceleration identification coefficient; In addition, the same driving habits (same accelerator / brake pedal amplitude and speed) can lead to completely different driving styles under different road conditions. The same driving operation may reflect an aggressive driving style in congested roads, but a normal driving style on unobstructed roads. Therefore, road condition recognition is the basis for driving style recognition.

[0016] The average speed and cruising time ratio are selected to characterize the working condition characteristics, and the average speed under the working condition segment speed sequence is u vehic ( t R)for: ; When a vehicle is running, large fluctuations in speed and steady speed passing are two completely different road conditions. It is insufficient to characterize the road condition characteristics by the cruising time ratio, so a parameter that can better reflect the current road condition information is needed. This scheme defines a speed fluctuation index to evaluate the smoothness-congestion degree of the road condition at this time, that is, the road condition passability evaluation coefficient. A generalized speed valley GSV is defined under a certain speed sequence, that is, the speed continues to decrease and then rises without stopping. The speed peak GSP is defined as the speed continues to increase and then begins to decrease. In order to avoid repeated calculations caused by small fluctuations in speed, a speed change value of 10km / h is defined as a GSP / GSV.

[0017] The number of GSP and GSV that appear in a unit recognition cycle is defined as the vehicle speed fluctuation (VelocityFluctuation, γ vfluc )as follows: ; In the formula, To identify the number of GSPs in a cycle, To identify the number of GSVs in the period; Speed ​​fluctuation γ vfluc The real-time calculation principle is as follows Figure 4 Take a certain speed sequence in the cycle condition as an example, Figure 4 As shown, by counting the number of velocity peaks , the number of speed valleys , substituting into the formula, we can get the vehicle speed fluctuation γ vfluc , showing the speed fluctuation of the speed sequence sample γ vfluc The real-time calculation result is 0.064.

[0018] In summary, the study a c R and A c R To characterize the driver's driving style, the driving style evaluation coefficient is defined as ECds (Driving Style Evaluation Coefficient). u vehic (t R ) and γ vflucTo characterize the real-time traffic style, the road trafficability evaluation coefficient ECrt (Evaluation Coefficient of Road Trafficability) is used.

[0019] EC ds The two inputs of the fuzzy controller are the acceleration identification coefficients ( a c R ) and impact degree identification coefficient ( A c R ), the output is the driving style evaluation coefficient EC ds .in a c R ∈[0, 0.5], A c R ∈[0, 0.5], EC ds ∈[0, 1].

[0020] EC rt The two inputs to the fuzzy controller are the average speed. u ( t R ) and vehicle speed fluctuation identification coefficient γ vfluc , the output is the road condition passability evaluation coefficient EC rt . Where utR∈[0,150], γvfluc∈[0,1], EC rt ∈[0, 1]. The design logic of the fuzzy rule base is: a c R and A c R The smaller, the more stable. EC ds The smaller;` u vehic (t R ) and γ vfluc The smaller, EC rt The smaller the value, the results are shown in Table 1 and Table 2: Table 1 ECds fuzzy rule base ; Table 2 ECrt fuzzy rule base ; The membership function and output surface of the fuzzy controller are as follows: Figure 4 and Figure 5 shown.

[0021] Considering the diverse travel needs, the recognition algorithm of a single scale window cannot meet the recognition needs. For example, when the road is congested and the vehicle speed is slow, it is unreasonable to choose the driving distance as the switching window. The time is too long and it is not reasonable. The time window recognition module should be selected. For smooth roads, the vehicle speed is fast, and it is not reasonable to choose the driving time as the switching window. The driving distance window recognition module should be selected. Define the control condition recognition cycle as T R : (Control Condition Recognition Time); define the control condition action cycle as T I (Control Condition Implementation Time).

[0022] Based on the parameter recognition algorithm of the variable scale recognition window, the key threshold parameters are optimized to achieve the accuracy and effectiveness of the recognition results. The specific recognition principle is as follows: ; In the formula, is the recognition period in time scale, is the identification cycle at the driving distance scale; Introducing control coefficient I m Establishing a control condition identification cycle T R and control implementation cycle T I The relationship between: T I = I m * T R ,in: ; In the formula, is the recognition period in time scale, is the identification cycle at the driving distance scale; Taking into account that the driving characteristics during driving have a certain "driving inertia", that is, they will not suddenly change between congested and intense driving and smooth and calm driving in a short period of time, the proportional conversion method used in the recognition module is simple in logic and can meet the control requirements.

[0023] Finally, the recognition cycle and control cycle of the recognition process based on the variable scale window are calculated as follows: ; ; The specific steps of the algorithm are as follows: (1) At any time t i Next, the acquisition time length is [( t i - T R s), t i ], thereby calculating the travel distance and using it in the distance scale information module; (2) Combining the actual time and distance information, and comparing it with the recognition period under the time scale and distance scale, the final recognition period is obtained; (3) Calculate the control period of action under time scale and distance scale T I , t i The time and travel distance after the moment, whichever reaches the threshold value first, is the action termination moment; (4) The recognition result is [ t i , ( t i + T I )] period, using travel characteristics information t i + T I After that, continue timing and start the next recognition cycle.

[0024] , , I m It is an important threshold parameter that determines the accuracy of the recognition results in the driving style and road condition feature recognition module, and a reasonable value is required. The specific optimization process is as follows: Parameter identification under multi-dimensional constraints and actual result center distance evaluation coefficient D EC To describe the optimization problem as the goal, it can be expressed as follows: ; In the formula, J min Represents the minimum comprehensive evaluation index of the control system and defines the evaluation coefficient D EC The objective function is to minimize the evaluation coefficient D EC Optimize the function for the target; The calculation formula of parameter identification and actual result center distance is: dF(distance function), calculated as follows: ; Evaluation Index D EC Represents a Tr The average center distance after comparing the result within the identification cycle with the result identified in the working condition segment. The evaluation coefficient of the center distance between parameter identification and actual results is D EC The calculation is as follows: ; Based on the BB-MOPSO algorithm, the parameter optimization experiment is carried out, and the average value of each parameter in each generation is selected to describe the change law of the optimization variable in the iteration process. The results are as follows Figures 6 to 9 shown.

[0025] It can be seen that after about 70 optimization iterations, each optimization variable finally converges to a stable value. The algorithm converges well, and the convergence trajectory is different. t R preset The fluctuation range is [70, 200]. After about 20 iterations, it stabilizes at [75, 85] and finally converges to 80s. t R [M t =M preset ] The fluctuation range is [0.4,1.5], and it gradually converges after 0.8km. After about 70 iterations, t R [M t =M preset ] Converges to around 0.6. I m The fluctuation range of is [0.05, 0.6] and converges to 0.25 after about 50 iterations. D isEC The fluctuation range is [0.3, 0.9], and the whole process is a "step-by-step" downward fluctuation. After 60 iterations, D isEC Converges to 0.3. Under the four cyclic road conditions, the parameter convergence curves are different. When the road is congested, the parameter value is small, the curve is relatively flat and the fluctuation is small, which means that the value is relatively sensitive. When the road is smooth, the curve fluctuates greatly. The final parameter value is [ t R preset , t R[M t =M preset ] , I m ] T =[0.9, 0.6, 0.25] T .

[0026] The final parameter value is [ t R preset , t R [ M t =M preset ]、 I m ] T =[0.9, 0.6, 0.25] T .

[0027] Standard deviation of shock intensity under four driving cycle conditions based on parameter identification results of variable scale identification window a c R and A c R The calculation results are as follows Fig.10 shown.

[0028] Driving style evaluation coefficient based on fuzzy algorithm EC ds and road condition passability evaluation coefficient EC rt The recognition result is as follows Fig.11 shown.

[0029] from Fig.10 and Fig.11 It can be seen that the driving style of working conditions changes frequently, and this frequent change may lead to frequent switching of various working modes; the changes in working condition types are relatively rare. The driving style and road condition feature recognition algorithm based on fuzzy algorithm can better complete the online recognition function and the recognition results are good.

[0030] Step 3: Input the driving style and road condition characteristic evaluation parameters acquired in real time into the "strategy switching" decision module for switching between the top-level energy modes CD-Blend and CD-EV+CS-Blend.

[0031] During the driving process of the whole vehicle, the theoretical SoC curve satisfies the SoC reaching the minimum limit at the end of the driving mileage to achieve efficient use of electric energy. The linear decline trajectory of SoC ignores the power maintenance and charging status to a certain extent, and the overall trend will appear as an approximate oblique line. From the perspective of the SoC linear shape, the design and optimization of the intelligent switching energy management strategy is the active planning and control of the timing and shape of the SoC curve for power consumption, stability and increase. On the other hand, it is the adjustment and optimization of the battery energy utilization method within a unit of driving mileage. The intelligent switching energy management strategy based on travel characteristics combines the APU system working mode optimization in Chapter 3 and the "approximately optimal" power reduction strategy designed in Chapter 5, by defining the parameters representing the "power mileage density". k cr To carry out the "strategy switching" of CD-EV, CS-Blend and CD-Blend, the APU system working mode is switched through the vehicle speed sequence characteristic information determined by driving style and road conditions. The offline test method is used to obtain the influence of the evaluation coefficient on the equivalent energy consumption, comprehensive emissions, power battery life attenuation rate and comprehensive evaluation index of the whole vehicle, and the multi-objective optimization and decision-making based on the optimal comprehensive evaluation index value form a rule control table to realize the "intelligent switching" function proposed in the study.

[0032] The research method and implementation process of the intelligent "mode switching" energy management strategy based on travel characteristics proposed in this scheme are as follows: Fig.12 As shown in the figure. The implementation process can be divided into four parts: (1) The representative driving cycle is divided into working condition segments and the data is pre-processed for noise reduction. The obtained vehicle speed sequence information is used to identify the driving style and working condition segment type. (2) Define and extract the characteristic parameters that characterize the driving style and working condition type, design the fuzzy recognition algorithm and the working condition sequence acquisition algorithm of the "variable scale window", and input the driving style and road condition characteristic evaluation parameters obtained in real time into the "strategy switching" decision module. (3) Based on the statistical analysis of the simulation test results, define the "intelligent mode switching" threshold value for the switching and decision of the top energy modes CD-Blend and CD-EV+CS-Blend. (4) Based on the offline test data, obtain the influence of the evaluation coefficient on the equivalent energy consumption of the whole vehicle, the comprehensive emission index, the power battery life attenuation rate and the comprehensive evaluation index. Based on the optimal comprehensive evaluation index, form the optimal strategy allocation rule base for implementing the "intelligent switching strategy" and the optimal power allocation method of the APU system.

[0033] 1. Define switching parameters k cr Combined with the real-time SoC size, the switching parameters of power reduction mode and power maintenance mode k cr Calculated as: ; In the formula, ∆SoC is the difference between theory and practice, k cr It can be considered to represent "power mileage density", which represents the distribution density of power over mileage and represents the distribution of battery power over the entire driving mileage.

[0034] 2. Determine k cr Threshold switching parameter Defining the comprehensive cost function F c , which includes three sub-costs: fuel cost, electricity cost and battery cost caused by battery capacity decay. Comprehensive cost function F c , the fuel cost function and the electricity cost function are calculated as follows: ; In the formula, , and They are fuel price (yuan / L), electricity price (yuan / kwh) and battery depreciation price (yuan / capacity). is the fuel cost, i.e. the amount of fuel consumed in this cycle (L), is the cost of electricity use, i.e., electricity (kwh), is the battery usage cost, that is, the battery capacity attenuation, m is the fuel consumption per unit time (g / s).

[0035] In this embodiment, the battery life is evaluated and the information is used by using the fitted empirical formula. The cost of power loss is the total cost of replacing the battery pack after the battery capacity decays by more than 30%, which is evenly allocated to the cost consumption during this driving process. The comprehensive cost function is used for the final decision as a compromise. Using more relatively cheap electricity will increase the battery capacity decay rate to a certain extent, but it will not increase significantly.

[0036] Set driving cycle conditions, vehicle weight, battery SoC 0 Value, vehicle estimated mileage M preset The accumulation is performed every 5 km, increasing from 100 km to 300 km, as shown in Table 3.

[0037] Table 3 Simulation test parameter settings ; The comprehensive cost function and the results of driving range under different driving ranges are as follows: Figures 13 to 16 As shown, we can see that under the same conditions (vehicle weight, SoC 0 There is an intersection point between the two energy strategy mode images under the conditions of , cycle conditions, and the statistical results under different factors are analyzed. As the mileage increases, F c Follow M preset Both show an increasing trend. The difference is that the increase slope is different. Under the CD-Blend strategy, F c Follow M preset Increase slowly, not very sensitive, showing a linear increase throughout the process; under the CD-EV+CS-Blend strategy, when M preset When smaller, F c Follow M preset Increase slowly, F c The function has a small slope, and then with the start-up of the APU, F c Follow M preset The increase is more obvious.

[0038] In order to increase the observation samples, the study expanded the vehicle weight in the simulation working conditions on this basis. m 、SoC0 settings are as follows: m =1500, 1550, 1600, 1650, 1700, 1750, 1800; SoC0=0.9, 085, 0.8, 0.75, 0.7, 0.65, 0.6, 0.55, 0.5. In this way, under the CD-Blend strategy and CD-EV+CS-Blend strategy, 7×9×4=252 F c Function intersections, these intersections are the "switch points" of the two strategies. These 252 intersections are statistically analyzed, such as Fig.17 shown.

[0039] from Fig.11 It can be seen that these 252 F c The function intersection points are distributed in a strip-shaped area, showing a relatively obvious "divergent distribution". Statistical laws show that these "turning points" correspond to k cr Most of the values ​​are in the range of [0.2, 0.25], so we can assume that k crThe threshold switching parameters of the value can be set within this range. After analysis and decision making, the least squares linear fitting k cr The value selected as the initial value of the threshold parameter is 0.21 (1 / 100km).

[0040] 3. Switching between top energy modes CD-Blend and CD-EV+CS-Blend In the estimated mileage M preset Based on real-time SoC identification, the real-time k update Value. k update ≤ k cr When , CD-EV+CS-Blend strategy is adopted; when k update > k cr When using CD-Blend mode.

[0041] Step 4: Evaluate parameters based on different driving styles and road conditions, and switch and decide on the APU system working mode.

[0042] Select the APU system working mode under the CD-Blend strategy and CD-EV+CS-Blend strategy N csop =1+ line , N csop =2, N csop =3 and N csop =4 Experimental research was carried out. Based on the test result data of 318 operating condition fragments formed under 4 driving cycle conditions, the control law was summarized by traversal optimization calibration. The results of strategy switching law are summarized as follows Fig.18 and Fig.19 As shown: Fig.18 and Fig.19In the figure, each small square contains relevant information about the operating condition segment, namely, 2 characteristic indicators and 4 performance indicators of the travel characteristics. Each small square is represented by a color under the corresponding performance indicator, and the depth of the color represents the distribution range of the indicator value under the travel characteristics corresponding to the small square. The result of each square in the figure is obtained by processing the mean of the test data. It can be seen that under the CD-Blend strategy, the preference for the two indicators of energy consumption and emissions (the square colors are mostly light), and the preference for comprehensive indicators (the square colors are mostly darker), the overall distribution trend law is roughly the same as the CD-EV+CS-Blend strategy. The strategy corresponding to the small square with the best comprehensive performance indicator is marked as the final "mode switching rule" based on travel characteristic recognition and "smart switching" decision, and is organized as follows: Fig. 20 shown.

[0043] It can be seen that under different travel characteristics, the APU system N csop =1+ line , N csop =2, N csop =3 and N csop =4 distribution switching in working mode. Under CD-EV+CS-Blend strategy, N csop =2, N csop =3 and N csop =4 The proportion is relatively even. In the case of aggressive driving style and congested traffic, APU is used. N csop =1+ line , N csop =2 working mode; APU is used when driving style is stable and road conditions are smooth. N csop =3, N csop =4 working mode; when the traffic is congested, the driving style is aggressive, the vehicle speed fluctuates greatly, and starts and stops frequently, the strategy adopts the CD-EV pure electric drive mode (SoC≥0.45).

[0044] The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described 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.

[0045] The above-mentioned embodiments only express several implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

[0046] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A range extender mode switching control method considering energy consumption, emission and battery life, characterized in that: The method comprises: Collect vehicle driving data in real time, perform noise reduction preprocessing on the vehicle speed sequence data through impulse noise filter and smooth curve filter, divide the working condition segments and extract characteristic parameters; Based on the vehicle speed sequence data and the preset mileage, the driving style evaluation coefficient is identified online based on the fuzzy algorithm and the variable scale window algorithm. EC ds and road condition passability evaluation coefficient EC rt , and optimize key threshold parameters; Defining the comprehensive cost function F c and switch parameters k cr , combined with real-time battery SoC and estimated driving range M preset , switch the top energy modes CD-EV+CS-Blend and CD-Blend; Dynamically switch the APU system working mode according to driving style and road conditions.

2. The method according to claim 1, characterized in that The impulse noise filter and the smoothing curve filter are defined as: Impulse Noise Filter: ； In the formula, represent The speed of the car at the moment, Represents the driving speed after filtering; Smooth Curve Filter: ； represent The driving speed after filtering at each moment, F Represents the vehicle speed weight.

3. The method according to claim 1, characterized in that The online identification driving style evaluation coefficient EC ds and road condition passability evaluation coefficient EC rt , and optimize key threshold parameters, including: Driver style recognition coefficient A c R , the evaluation indicators are defined as follows: ； In the formula, SD A c R Represents the standard deviation of the impact degree within a working condition identification cycle, A c R Represents the average absolute value of the typical working condition impact degree of the working condition category of ordinary drivers in a working condition identification cycle, T R represents the length of the recognition cycle, A c R ( t R )represent t R The impact degree identification coefficient at the moment, SD A c R Reflects the discrete degree of impact within any time period. a c R represents the acceleration identification coefficient, Indicates the degree of impact; The average speed and cruising time ratio are selected to characterize the working condition characteristics, and the average speed under the working condition segment speed sequence is u vehic ( t R )for: ； Define the speed fluctuation index to evaluate the smoothness and congestion of the current road conditions and obtain the road condition passability evaluation coefficient; The number of speed peaks GSP and speed valleys GSV that appear within a unit recognition period is defined as the vehicle speed fluctuation degree. γ vfluc as follows: ； In the formula, To identify the number of GSPs in a cycle, To identify the number of GSVs in the period; The acceleration identification factor a c R and impact degree identification coefficient A c R As the input of the fuzzy controller, the output driving style evaluation coefficient EC ds ,in a c R ∈[0,0.5], A c R ∈[0,0.5], EC ds ∈[0,1]; The average speed u ( t R ) and vehicle speed fluctuation identification coefficient γ vfluc As the input of the fuzzy controller, the output road condition passability evaluation coefficient EC rt ,in u ( t R )∈[0,150],γ vfluc ∈[0,1], EC rt ∈[0,1].

4. The method according to claim 3, characterized in that The identification period of the variable scale window algorithm T R , specifically: ； In the formula, is the recognition period in time scale, is the identification cycle at the driving distance scale; Introducing control coefficient I m Establishing a control condition identification cycle T R and control implementation cycle T I The relationship between: T I = I m * T R ,in: ； In the formula, is the recognition period in time scale, is the identification cycle at the driving distance scale; Based on the variable scale window recognition process, the recognition cycle and control cycle are calculated as follows: ； 。 5. The method according to claim 4, characterized in that The key threshold parameters are optimized, and the specific optimization process is as follows: Parameter identification under multi-dimensional constraints and actual result center distance evaluation coefficient D EC The optimization problem is described as follows: ； In the formula, J min Represents the minimum comprehensive evaluation index of the control system and defines the evaluation coefficient D EC The objective function is to minimize the evaluation coefficient D EC Optimize the function for the target; The calculation formula of parameter identification and actual result center distance is: dF The calculation is as follows: ； Evaluation coefficient D EC Represents a T R The average value of the center distance after comparing the result within the identification cycle with the result identified in the working condition segment, and the evaluation coefficient of the center distance between the parameter identification and the actual result D EC The calculation is as follows: ； By iterating D isEC Converge and select final parameters The value of .

6. The method according to claim 1, characterized in that The comprehensive cost function F c Includes: Fuel usage costs , Electricity usage cost and battery usage costs , specifically: ； In the formula, , and They are fuel price, electricity price and battery depreciation price respectively. is the fuel consumption per unit time, Indicates the required power, The duration of the power consumption hybrid mode, The duration of pure electric mode is the power consumption. is the conversion factor of oil and electricity prices; The switching parameters k cr Combined with real-time battery SoC sizing calculation: ； In the formula, ∆SoC is the difference between theory and practice, k cr It is a switching parameter, which stands for power-mileage density and represents the distribution density of power over mileage.

7. The method according to claim 5, characterized in that The switching of the top energy modes CD-EV+CS-Blend and CD-Blend is specifically as follows: In the estimated mileage M preset Based on the real-time SoC identification, the real-time switching parameters are calculated k update value, k update Indicates the real-time update value of the power mileage density, that is, the real-time switching parameter value. k update ≤ k cr When , CD-EV+CS-Blend strategy is adopted; when k update > k cr When using CD-Blend mode.

Images