Scene working condition construction method and energy management system based on same

Abstract

The invention provides a scene working condition construction method and a vehicle energy management system based on the method, and the method comprises the steps: S1, collecting the real vehicle working conditions of vehicles in a certain park scene, and constructing a real vehicle working condition database; s2, on the basis of the real vehicle working condition database, constructing a working condition state recognition method based on micro-stroke fragments, and generating a working condition state sequence library by recognizing the states of the micro-stroke working condition fragments; s3, establishing a frequent transition sequence mining algorithm based on a historical state sequence, processing the working condition state sequence library, and extracting a working condition state sequence which frequently appears in the real vehicle working condition database; and S4, in the real vehicle working condition database, extracting working condition fragments corresponding to the frequently appearing working condition state sequence, and splicing the working condition fragments to form a scene working condition. According to the method, the scene working condition of the new energy vehicle in the park application scene is constructed, so that the scene working condition capable of representing the kinematics characteristics and the dynamics characteristics of the new energy vehicle in the application scene is constructed.

Description

Technical Field

[0001] The present invention relates to the technical field of automobile energy-saving control, and in particular to a scenario operating condition construction method and an energy management system suitable for new energy vehicles based on the method. Background Art

[0002] With the rapid economic development, industrial parks are booming. Creating low-carbon and energy-saving green industrial parks has become an effective way to alleviate energy shortages. The number of industrial parks and the number of special vehicles in them have increased year by year, resulting in an increase in energy consumption in the parks, especially in the transportation sector. Therefore, electrification of special vehicles in industrial parks and reducing energy consumption have become key technical means.

[0003] The vehicle energy management strategy (EMS) is the core of energy consumption optimization for new energy vehicles. The EMS of new energy vehicles is often calibrated based on a large number of standard operating condition tests. The calibration strategy is stored in the controller in the form of a rule table related to the operating conditions, vehicle status, etc. When the car is driving, the real-time control quantity is solved according to the rule table and the actual vehicle operating conditions. In order to make the calibrated EMS fit the vehicle driving conditions to achieve higher economy, the calibration conditions must comprehensively and accurately cover the kinematic and time series characteristics of the actual driving scenarios of the vehicle. For vehicles driving in the park, the driving condition scenario characteristics are obvious. It is extremely necessary to construct scenario conditions that can cover these characteristics to improve the calibration efficiency and energy saving of EMS for special vehicles in the park.

[0004] However, the current calibration conditions mostly use standard conditions, such as the Worldwide Harmonized Light Vehicle Test Cycle (WLTC) and the China Light Vehicle Driving Conditions (CLTC). These standard conditions are difficult to fully and accurately represent the application scenarios of the park. The maximum speeds of WLTC and CLTC are about 130km / h and 90km / h, respectively, which far exceed the speed limit requirements of the park. And their idling ratios are only about 13% and 20%, respectively, but the speed limit in the park is low, the flow of people is large, and the idle ratio of the shuttle bus is relatively high. Therefore, the standard conditions cannot reflect the driving conditions of the park, and the EMS based on its calibration is difficult to achieve energy saving of park vehicles.

[0005] The existing mainstream scenario working condition construction method has the following process: first, the working conditions in the actual vehicle working condition database are cut into short-stroke segments; then the short-stroke segments are identified and numbered; then the Markov-Monte Carlo principle (MMC) is used to splice short-stroke segments with different state numbers; finally, the corresponding state working condition segments are extracted from the database to form scenario working conditions. However, there are problems with this method. The short-stroke segments start and end with the idle segment, contain multiple driving mode working condition segments, and can be further cut into micro-stroke segments. When solving the working condition characteristics of short-stroke segments, the characteristics of micro-stroke segments with different modes or large feature differences are easily neutralized, so that the solution results cannot reflect the real characteristics, and it is difficult to construct a scenario working condition that reflects the real kinematic characteristics of the scene. In addition, MMC is highly random and cannot find the working condition state sequence that frequently appears in the scenario database. The scene time series characteristics cannot be considered when constructing the working condition, resulting in the inability of the existing method to construct a scenario working condition that can reflect the real time series characteristics of the scene, and the calibrated EMS is difficult to adapt to the application scenario.

[0006] Through searching patent documents, it was found that the invention patent with publication number CN107463992B discloses a method for predicting the driving condition of a hybrid vehicle based on segment waveform training. The method includes: dividing the entire vehicle driving condition into a combination of several driving condition segments, wherein one driving condition segment refers to the process from vehicle start to the first parking brake, including the starting phase, driving phase, and braking phase; using the driving condition segments to train the neural network, and then through the powerful fitting ability of the neural network, the input driving condition history information is analyzed and calculated, and the historical information is matched with the training driving condition segments; in the training driving condition segments, the part that is closest to the historical information is found, and the following condition information of this part is output as the prediction result to complete the prediction of future driving conditions. This patent focuses on the prediction of vehicle driving conditions, and does not discuss specific scenario conditions and related energy management in depth.

[0007] In summary, in view of the above-mentioned problems of the prior art, studying a scenario condition construction method and an energy management system suitable for new energy vehicles based on the method has become a key task that needs to be solved urgently. Summary of the invention

[0008] In view of the defects in the prior art, an object of the present invention is to provide a scenario condition construction method and an energy management system suitable for new energy vehicles based on the method.

[0009] A method for constructing a scene condition according to the present invention includes:

[0010] Step S1, collecting actual vehicle operating conditions in a certain park scene and building an actual vehicle operating condition database;

[0011] Step S2, based on the actual vehicle working condition database, constructing a working condition state recognition method based on micro-stroke segments, and generating a working condition state sequence library by identifying the states of the micro-stroke working condition segments;

[0012] Step S3, establishing a frequent transition sequence mining algorithm based on historical state sequences, processing the operating state sequence library, and extracting the operating state sequences that frequently appear in the actual vehicle operating condition database;

[0013] Step S4: extracting operating condition segments corresponding to frequently occurring operating condition state sequences from the actual vehicle operating condition database, and splicing them to form scenario operating conditions.

[0014] Preferably, step S2 includes the following sub-steps:

[0015] Step S2.1, based on the actual vehicle operating condition database, cutting the actual vehicle operating condition into short-stroke segments;

[0016] Step S2.2, cutting each short-trip segment into micro-trip segments according to different driving modes;

[0017] Step S2.3, performing kmeans clustering on the micro-trip segments under different driving modes, and marking the category number of each micro-trip segment as the working condition;

[0018] Step S2.4, based on the micro-stroke segments after the marked category number, the micro-stroke segments of each working condition in the actual vehicle working condition database are state-marked, that is, the actual vehicle working condition sequence is converted into a working condition state sequence to obtain a working condition state sequence library.

[0019] Preferably, in step S2.1, the short trip segment is a working condition segment in which both the starting segment and the ending segment are idle segments; the idle segment is a segment in which the real-time vehicle speed is less than or equal to 4 km / h and the absolute value of the acceleration is less than or equal to 0.15 m / s 2 The speed sequence.

[0020] Preferably, in step S2.2, the driving mode classification standard is as follows: the acceleration mode is when the real-time vehicle speed is greater than 4 km / h and the real-time acceleration is greater than 0.15 m / s 2 The vehicle speed sequence is cut according to the acceleration mode to obtain the acceleration micro-stroke segment; the deceleration mode is when the real-time vehicle speed is greater than 4km / h and the real-time acceleration is less than -0.15m / s 2 The speed sequence is cut according to the deceleration mode to obtain the deceleration micro-stroke segment; the cruise mode is when the real-time speed is greater than 4km / s and the absolute value of the real-time acceleration is less than 0.15m / s 2 The speed sequence is cut according to the cruise mode to obtain the cruise micro-trip segment; the idle mode is when the real-time vehicle speed is less than or equal to 4km / h and the absolute value of acceleration is less than or equal to 0.15m / s2 The vehicle speed sequence is cut according to the idle mode to obtain the idle micro-stroke fragments.

[0021] Preferably, in step S2.3, the acceleration micro-stroke segments are divided into three categories according to the acceleration time, initial velocity and average velocity characteristics; the deceleration micro-stroke segments are divided into four categories according to the deceleration time, final velocity and average velocity characteristics; and the cruising micro-stroke segments are divided into three categories according to the average velocity and cruising time characteristics.

[0022] Preferably, in step S2.3, the operating conditions of the acceleration micro-stroke segments are marked as 1, 2, and 3 respectively; the operating conditions of the deceleration micro-stroke segments are marked as 4, 5, 6, and 7 respectively; the operating conditions of the cruising micro-stroke segments are marked as 8, 9, and 10 respectively; and the operating condition of the idling micro-stroke segments is marked as 11.

[0023] Preferably, step S3 includes the following sub-steps:

[0024] Step S3.1, for each short-stroke segment in the operating state sequence library, determine the starting segment, record the operating state of the starting segment, set the idle micro-stroke segment as the starting segment of the short-stroke segment, and mark the operating state as 11;

[0025] Step S3.2, based on the operating state sequence library, calculate the transition probability between each two states of the 11 operating states determined in step S2.3, as shown in the following formula:

[0026]

[0027] Among them, p i,j represents the transition probability from operating state i to operating state j; z i,j Represents the number of times the transition from operating state i to operating state j appears in the operating state sequence library;

[0028] Step S3.3, for any operating state i among the 11 operating states, arrange the transition probabilities of the operating state i to each of the remaining operating states from large to small;

[0029] Step S3.4, starting from the working state 11, find the next working state that the working state 11 can easily transfer to in the working state sequence library, the next working state that can easily transfer to is the working state with a larger transfer probability calculated according to step S3.3, record the working state 11 as the mth level historical state sequence, and record the next working state that can easily transfer to as the nth level easy-to-transfer state;

[0030] Step S3.5, adding the nth level easy-to-transfer state to the end of the mth level historical state sequence to form a new historical state sequence, and the historical state sequence is defined as the m+1th level historical state sequence;

[0031] Step S3.6, find the next level of working condition that the m+1 level historical state sequence can easily transfer to in the working condition sequence library, and construct the next level of historical state sequence;

[0032] Step S3.7, repeat steps S3.5 to S3.6 until the operating state corresponding to the idle micro-stroke segment is found, that is, stop at operating state 11.

[0033] The present invention also provides an energy management system suitable for new energy vehicles, based on the above-mentioned scenario condition construction method, comprising:

[0034] Module M1 collects the actual vehicle operating conditions in a certain park scenario and builds a database of actual vehicle operating conditions;

[0035] Module M2, based on the actual vehicle working condition database, constructs a working condition state recognition method based on micro-stroke fragments, and generates a working condition state sequence library by identifying the state of micro-stroke working condition fragments;

[0036] Module M3, establishes a frequent transition sequence mining algorithm based on historical state sequences, processes the operating state sequence library, and extracts the operating state sequences that frequently appear in the actual vehicle operating condition database;

[0037] Module M4 extracts the operating condition fragments corresponding to the frequently occurring operating condition state sequences from the actual vehicle operating condition database and splices them to form a scenario operating condition;

[0038] Module M5, based on scenario conditions, uses rule learning algorithms to calibrate energy management strategies, outputs calibrated rule tables, and builds an energy management system suitable for new energy vehicles in application scenarios.

[0039] Preferably, module M5 comprises:

[0040] Unit M5.1, based on a fuel cell hybrid vehicle, uses a dynamic programming algorithm to solve the scenario conditions and obtain the optimal control law, which includes the SOC trajectory, fuel cell output power and required power; the combination of the SOC trajectory value, fuel cell output power value and required power value at each moment is regarded as a sample;

[0041] Unit M5.2, based on samples, uses rule learning algorithm to learn the relationship between fuel cell output power, required power and SOC.

[0042] Preferably, unit M5.2 comprises the following subunits:

[0043] Subunit M5.2.1, clustering the samples using a subtractive clustering algorithm to obtain the classification numbers of SOC, fuel cell output power, and required power;

[0044] Subunit M5.2.2, based on the number of classifications, sets the SOC and the required power as the input of the energy management strategy, sets the output power of the fuel cell as the output, and for each input combination formed by the SOC and the required power, finds the corresponding fuel cell output power type with the highest occurrence frequency, and forms a rule table of the type;

[0045] Subunit M5.2.3, based on the number of classifications, solves the cluster center of each category, fills the value of the cluster center into the corresponding type in the rule table, and forms the calibration rules of the energy management strategy.

[0046] Compared with the prior art, the present invention has the following beneficial effects:

[0047] 1. The present invention realizes the construction of scenario conditions for new energy vehicles in park application scenarios, thereby realizing the construction of a scenario condition that can represent the kinematic characteristics and dynamic characteristics of new energy vehicles in application scenarios.

[0048] 2. The scene condition construction algorithm provided by the present invention proposes a state recognition method based on micro-condition fragments, which can retain the significant kinematic characteristics of the scene condition.

[0049] 3. The present invention also takes into account the state transition probability in the construction of scene conditions and realizes the mining of frequent state transition sequences in the database, thereby retaining the time series characteristics in the database. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings:

[0051] Figure 1 A schematic diagram of a flow chart of a method for constructing a scene condition in an embodiment of the present invention;

[0052] Figure 2 It is a principle diagram of a frequent transition sequence mining algorithm based on historical state sequences in an embodiment of the present invention. DETAILED DESCRIPTION

[0053] The present invention is described in detail below in conjunction with specific embodiments. The following embodiments will help those skilled in the art to further understand the present invention, but are not intended to limit the present invention in any form. It should be noted that, for those of ordinary skill in the art, several changes and improvements can also be made without departing from the concept of the present invention. These all belong to the protection scope of the present invention.

[0054] The present invention discloses a scenario condition construction method suitable for new energy vehicles based on micro-granularity feature recognition and frequent transfer sequence mining, and develops an energy management system suitable for new energy vehicles based on this method.

[0055] Embodiment 1:

[0056] Figure 1 The figure is a flow chart of a method for constructing a scene condition in an embodiment of the present invention.

[0057] like Figure 1 As shown, this embodiment provides a method for constructing a scene condition, comprising the following steps:

[0058] Step S1, collecting actual vehicle operating conditions in a certain park scenario and building an actual vehicle operating condition database.

[0059] Step S2, based on the actual vehicle working condition database, construct a working condition state recognition method based on micro-stroke segments, and generate a working condition state sequence library by identifying the states of micro-stroke working condition segments.

[0060] Specifically, step S2 includes the following sub-steps:

[0061] Step S2.1, based on the actual vehicle operating condition database, cutting the actual vehicle operating condition into short-stroke segments.

[0062] In this embodiment, the short trip segment is a working condition segment in which both the starting segment and the ending segment are idle segments; the idle segment is a segment in which the real-time vehicle speed is less than or equal to 4km / h and the absolute value of the acceleration is less than or equal to 0.15m / s 2 The speed sequence.

[0063] Step S2.2, cutting each short-trip segment into micro-trip segments according to different driving modes.

[0064] In this embodiment, the driving mode classification criteria are as follows: the acceleration mode is when the real-time vehicle speed is greater than 4 km / h and the real-time acceleration is greater than 0.15 m / s 2 The vehicle speed sequence is cut according to the acceleration mode to obtain the acceleration micro-stroke segment; the deceleration mode is when the real-time vehicle speed is greater than 4km / h and the real-time acceleration is less than -0.15m / s 2 The speed sequence is cut according to the deceleration mode to obtain the deceleration micro-stroke segment; the cruise mode is when the real-time speed is greater than 4km / h and the absolute value of the real-time acceleration is less than 0.15m / s 2 The speed sequence is cut according to the cruise mode to obtain the cruise micro-trip segment; the idle mode is when the real-time vehicle speed is less than or equal to 4km / h and the absolute value of acceleration is less than or equal to 0.15m / s 2 The vehicle speed sequence is cut according to the idle mode to obtain the idle micro-stroke fragments.

[0065] In step S2.3, kmeans clustering is performed on the micro-trip segments under different driving modes, and the category number of each micro-trip segment is marked as the working condition status.

[0066] Specifically, the acceleration micro-stroke segments are divided into three categories according to the acceleration time, initial velocity and average velocity characteristics; the deceleration micro-stroke segments are divided into four categories according to the deceleration time, final velocity and average velocity characteristics; the cruising micro-stroke segments are divided into three categories according to the average velocity and cruising time characteristics.

[0067] In this embodiment, the operating conditions of the acceleration micro-stroke segments are marked as 1, 2, and 3 respectively; the operating conditions of the deceleration micro-stroke segments are marked as 4, 5, 6, and 7 respectively; the operating conditions of the cruising micro-stroke segments are marked as 8, 9, and 10 respectively; and the operating condition of the idling micro-stroke segment is marked as 11.

[0068] Step S2.4, based on the micro-stroke segments after the marked category number, the micro-stroke segments of each working condition in the actual vehicle working condition database are state-marked, that is, the actual vehicle working condition sequence is converted into a working condition state sequence to obtain a working condition state sequence library.

[0069] Step S3, establishing a frequent transition sequence mining algorithm based on the historical state sequence, processing the operating state sequence library, and extracting the operating state sequences that frequently appear in the actual vehicle operating condition database.

[0070] Figure 2 It is a principle diagram of a frequent transition sequence mining algorithm based on historical state sequences in an embodiment of the present invention.

[0071] like Figure 2 As shown, step S3 includes the following sub-steps:

[0072] Step S3.1, for each short-stroke segment in the operating state sequence library, determine the starting micro-stroke segment, and record the operating state of the starting micro-stroke segment, set the idle micro-stroke segment as the starting segment of the short-stroke segment, and mark the operating state as 11.

[0073] Step S3.2, based on the operating state sequence library, calculate the transition probability between each two states in the 11 operating states determined in step S2.3, as shown in the following formula:

[0074]

[0075] Among them, p i,j represents the transition probability from operating state i to operating state j; z i,j Represents the number of times the transition from operating state i to operating state j occurs in the operating state sequence library.

[0076] Step S3.3, for any operating state i among the 11 operating states, arrange the transition probabilities of operating state i to each of the remaining operating states from large to small.

[0077] Step S3.4, starting from operating state 11, find the next operating state that operating state 11 can easily transition to in the operating state sequence library. The next operating state that is easy to transition to is the operating state with a larger transition probability calculated according to step S3.3. Operating state 11 is recorded as the mth level historical state sequence, and the next operating state that is easy to transition to is recorded as the nth level easy-to-transition state.

[0078] In this embodiment, the criterion for determining "easy to transfer into" is to select the top two operating conditions with the highest transfer probability; m=1, n=1.

[0079] Step S3.5, adding the nth level easy-to-transfer state to the end of the mth level historical state sequence to form a new historical state sequence, and the historical state sequence is defined as the m+1th level historical state sequence.

[0080] Step S3.6, find the next level of operating state that the m+1 level historical state sequence can easily transfer to in the operating state sequence library, and construct the next level of historical state sequence.

[0081] Step S3.7, repeat steps S3.5 to S3.6 until the operating state corresponding to the idle micro-stroke segment is found, that is, stop at operating state 11.

[0082] Step S4: extracting operating condition segments corresponding to frequently occurring operating condition state sequences from the actual vehicle operating condition database, and splicing them to form scenario operating conditions.

[0083] Embodiment 2:

[0084] This embodiment provides an energy management system applicable to new energy vehicles, based on a scenario condition construction method in the above embodiment 1, including:

[0085] Module M1 collects the actual vehicle operating conditions in a certain park scenario and builds a database of actual vehicle operating conditions.

[0086] Module M2, based on the actual vehicle working condition database, constructs a working condition state recognition method based on micro-stroke fragments, and generates a working condition state sequence library by identifying the state of micro-stroke working condition fragments.

[0087] Module M3 establishes a frequent transition sequence mining algorithm based on historical state sequences, processes the operating condition sequence library, and extracts the operating condition sequence that frequently appears in the actual vehicle operating condition database.

[0088] Module M4 extracts the operating condition fragments corresponding to the frequently occurring operating condition state sequences from the actual vehicle operating condition database and splices them into scenario operating conditions.

[0089] Module M5, based on scenario conditions, uses rule learning algorithms to calibrate energy management strategies, outputs calibrated rule tables, and builds an energy management system suitable for new energy vehicles in application scenarios.

[0090] Specifically, module M5 includes:

[0091] Unit M5.1, based on a fuel cell hybrid vehicle, uses a dynamic programming algorithm to solve the scenario conditions and obtain the optimal control law. The optimal control law includes the SOC trajectory, fuel cell output power and required power; the combination of the SOC trajectory value, fuel cell output power value and required power value at each moment is taken as a sample.

[0092] Unit M5.2, based on samples, uses rule learning algorithm to learn the relationship between fuel cell output power, required power and SOC.

[0093] Furthermore, unit M5.2 includes the following subunits:

[0094] Subunit M5.2.1, clustering the samples using a subtractive clustering algorithm to obtain the classification numbers of SOC, fuel cell output power and required power.

[0095] Subunit M5.2.2, based on the number of classifications, sets SOC and required power as inputs of the energy management strategy, sets the output power of the fuel cell as output, and for each input combination formed by SOC and required power, finds the corresponding fuel cell output power type with the highest frequency of occurrence to form a rule table of the type.

[0096] Subunit M5.2.3, based on the number of classifications, solves the cluster center of each category, fills the value of the cluster center into the corresponding type in the rule table, and forms the calibration rules of the energy management strategy.

[0097] Those skilled in the art know that, in addition to realizing the system and its various devices, modules, and units provided by the present invention in a purely computer-readable program code, it is entirely possible to realize the same functions in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers by logically programming the method steps. Therefore, the system and its various devices, modules, and units provided by the present invention can be considered as a hardware component, and the devices, modules, and units included therein for realizing various functions can also be regarded as structures within the hardware component; the devices, modules, and units for realizing various functions can also be regarded as both software modules for realizing the method and structures within the hardware component.

[0098] The above describes the specific embodiments of the present invention. It should be understood that the present invention is not limited to the above specific embodiments, and those skilled in the art can make various changes or modifications within the scope of the claims, which does not affect the essence of the present invention. In the absence of conflict, the embodiments of the present application and the features in the embodiments can be combined with each other arbitrarily.

Claims

1. A method for constructing a scene condition, characterized in that: The steps include: Step S1, collecting actual vehicle operating conditions in a certain park scene and building an actual vehicle operating condition database; Step S2, based on the actual vehicle operating condition database, constructing an operating condition state recognition method based on micro-stroke segments, and generating an operating condition state sequence library by identifying the states of micro-stroke operating condition segments; Step S3, establishing a frequent transition sequence mining algorithm based on historical state sequences, processing the operating state sequence library, and extracting operating state sequences that frequently appear in the actual vehicle operating condition database; Step S4, extracting the operating condition segments corresponding to the frequently occurring operating condition state sequence from the actual vehicle operating condition database, and splicing them to form a scene operating condition.

2. A method for constructing a scene condition according to claim 1, characterized in that: The step S2 includes the following sub-steps: Step S2.1, based on the actual vehicle operating condition database, cutting the actual vehicle operating condition into short-stroke segments; Step S2.2, cutting each short-trip segment into micro-trip segments according to different driving modes; Step S2.3, performing kmeans clustering on the micro-trip segments under different driving modes, and marking the category number of each micro-trip segment as the working condition; Step S2.4, based on the micro-stroke segments after the marked category number, the micro-stroke segments of each working condition in the actual vehicle working condition database are state-marked, that is, the actual vehicle working condition sequence is converted into a working condition state sequence to obtain a working condition state sequence library.

3. A method for constructing a scene condition according to claim 2, characterized in that: In step S2.1, the short trip segment is a working condition segment in which both the starting segment and the ending segment are idle segments; the idle segment is a segment in which the real-time vehicle speed is less than or equal to 4 km / h and the absolute value of the acceleration is less than or equal to 0.15 m / s 2 The speed sequence.

4. A method for constructing a scene condition according to claim 2, characterized in that: In step S2.2, the driving mode classification standard is as follows: the acceleration mode is when the real-time vehicle speed is greater than 4km / h and the real-time acceleration is greater than 0.15m / s 2 The vehicle speed sequence is cut into the actual vehicle working condition according to the acceleration mode to obtain acceleration micro-stroke segments; The deceleration mode is when the real-time vehicle speed is greater than 4km / h and the real-time acceleration is less than -0.15m / s 2 The vehicle speed sequence is cut into the actual vehicle working condition according to the deceleration mode to obtain a deceleration micro-stroke segment; The cruise mode is when the real-time vehicle speed is greater than 4km / h and the absolute value of the real-time acceleration is less than 0.15m / s 2 The vehicle speed sequence is cut into the actual vehicle working condition according to the cruise mode to obtain cruise micro-stroke segments; The idle mode is when the real-time vehicle speed is less than or equal to 4km / h and the absolute value of acceleration is less than or equal to 0.15m / s 2 The vehicle speed sequence is cut according to the idle mode to obtain the idle micro-stroke segment.

5. A method for constructing a scene condition according to claim 2, characterized in that: In the step S2.3, the acceleration micro-stroke segments are divided into three categories according to the acceleration time, initial velocity and average velocity characteristics; the deceleration micro-stroke segments are divided into four categories according to the deceleration time, final velocity and average velocity characteristics; and the cruising micro-stroke segments are divided into three categories according to the average velocity and cruising time characteristics.

6. A method for constructing a scene condition according to claim 5, characterized in that: In step S2.3, the operating conditions of the acceleration micro-stroke segment are marked as 1, 2, and 3 respectively; the operating conditions of the deceleration micro-stroke segment are marked as 4, 5, 6, and 7 respectively; the operating conditions of the cruising micro-stroke segment are marked as 8, 9, and 10 respectively; and the operating condition of the idling micro-stroke segment is marked as 11.

7. A method for constructing a scene condition according to claim 1, characterized in that: The step S3 includes the following sub-steps: Step S3.1, for each short-stroke segment in the operating state sequence library, determine the starting segment, record the operating state of the starting segment, set the idle micro-stroke segment as the starting segment of the short-stroke segment, and mark the operating state as 11; Step S3.2, according to the operating state sequence library, calculate the transition probability between every two states of the 11 operating states determined in step S2.3, as shown in the following formula: Among them, p i,j represents the transition probability from operating state i to operating state j; z i,j Represents the number of times the transition from operating state i to operating state j appears in the operating state sequence library; Step S3.3, for any operating state i among the 11 operating states, arrange the transition probabilities of the operating state i to each of the remaining operating states from large to small; Step S3.4, starting from the operating state 11, find the next operating state that the operating state 11 can easily transfer to in the operating state sequence library, the next operating state that can easily transfer to is the operating state with a larger transfer probability calculated according to step S3.3, record the operating state 11 as the mth level historical state sequence, and record the next operating state that can easily transfer to as the nth level easy-to-transfer state; Step S3.5, adding the nth level easy-to-transfer state to the end of the mth level historical state sequence to form a new historical state sequence, and the historical state sequence is defined as the m+1th level historical state sequence; Step S3.6, finding the next level of operating state that the m+1 level historical state sequence can easily transfer to in the operating state sequence library, and constructing the next level of historical state sequence; Step S3.7, repeat steps S3.5 to S3.6 until the operating state corresponding to the idle micro-stroke segment is found, that is, stop at operating state 11.

8. An energy management system suitable for new energy vehicles, based on a scenario construction method according to any one of claims 1 to 7, characterized in that: include: Module M1 collects the actual vehicle operating conditions in a certain park scenario and builds a database of actual vehicle operating conditions; Module M2, based on the actual vehicle working condition database, constructs a working condition state recognition method based on micro-stroke segments, and generates a working condition state sequence library by identifying the states of micro-stroke working condition segments; Module M3, establishing a frequent transition sequence mining algorithm based on historical state sequences, processing the operating state sequence library, and extracting the operating state sequences that frequently appear in the actual vehicle operating condition database; Module M4, extracting the operating condition fragments corresponding to the frequently occurring operating condition state sequence from the actual vehicle operating condition database, and splicing them to form a scene operating condition; Module M5, based on the scenario conditions, uses a rule learning algorithm to calibrate the energy management strategy, outputs a calibrated rule table, and constructs an energy management system suitable for new energy vehicles in the application scenario.

9. The energy management system suitable for new energy vehicles according to claim 8, characterized in that: The module M5 comprises: Unit M5.1, based on a fuel cell hybrid vehicle, solve the scenario conditions using a dynamic programming algorithm to obtain an optimal control law, which includes the SOC trajectory, fuel cell output power and required power; the combination of the SOC trajectory value, fuel cell output power value and required power value at each moment is taken as a sample; Unit M5.2, based on the samples, uses a rule learning algorithm to learn the relationship between the fuel cell output power, the required power and the SOC.

10. The energy management system suitable for new energy vehicles according to claim 9, characterized in that: The unit M5.2 includes the following subunits: Subunit M5.2.1, clustering the samples using a subtractive clustering algorithm to obtain the classification numbers of the SOC, the fuel cell output power and the required power; Subunit M5.2.2, based on the number of classifications, setting the SOC and the required power as inputs of the energy management strategy, setting the output power of the fuel cell as output, and for each input combination formed by the SOC and the required power, finding the corresponding fuel cell output power type with the highest occurrence frequency, and forming a rule table of types; Subunit M5.2.3, based on the classification number, solves the cluster center of each category, fills the value of the cluster center into the corresponding type in the rule table, and forms the calibration rule of the energy management strategy.

Images