Public vehicle operation management method based on big data analysis and related equipment

Abstract

The invention discloses a public vehicle operation management method based on big data analysis and related equipment, and the method comprises the steps: receiving a battery alarm signal of an electric public vehicle, and obtaining a battery parameter and vehicle operation data; inputting the current environment data and the real-time road condition data into a big data model to obtain a current operation scene of the vehicle; inquiring to obtain a battery parameter interval corresponding to the current operation scene of the vehicle; under the condition that the battery parameters are located in the battery parameter interval, vehicle energy consumption data in the preset kilometer interval are obtained, and according to the battery parameters, the vehicle energy consumption data and the current operation scene of the vehicle, the energy consumption trend of the electric public vehicle in the future preset time period is predicted; generating a power output mode of the electric public vehicle according to the energy consumption trend, the vehicle operation data and the driving behavior data; generating a brake adjusting instruction according to the power output mode; and the brake adjusting instruction is sent to the vehicle terminal, so that the vehicle terminal executes a brake strategy.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of electric public vehicles, and in particular to a public vehicle operation management method and related equipment based on big data analysis. Background Art

[0002] In modern urban transportation systems, electric public vehicles (such as electric buses) have become an important means of transportation. Electric public vehicles are not only environmentally friendly, but also can effectively reduce urban air pollution. However, the operation and management of electric public vehicles face many challenges, especially in battery management. Batteries are the core components of electric vehicles, and their performance directly affects the vehicle's mileage, safety, and operational efficiency. Therefore, how to effectively monitor and manage the battery status to ensure that the vehicle can operate stably under various operating conditions has become an urgent problem to be solved.

[0003] Traditional vehicle operation methods mainly rely on manual monitoring and experience judgment. In this mode, the battery status of the vehicle is usually judged by the driver or operation manager through simple instrument readings. When the battery alarm signal appears, the usual practice is to stop the vehicle immediately for inspection, or judge whether the battery needs to be replaced based on experience. Although this method is simple and direct, there are many problems in practical application.

[0004] First, traditional battery monitoring methods lack accuracy. Battery alarm signals are often based on simple parameters such as voltage, current and temperature, which are easily affected by environmental factors such as temperature changes and road conditions. Therefore, battery alarm signals sometimes give false alarms, that is, there is no real problem with the battery, but the system misjudges it due to some external factors. Such false alarms not only increase operating costs, but may also cause unnecessary vehicle downtime, affecting normal operations.

[0005] Secondly, in traditional vehicle operation methods, due to the lack of comprehensive monitoring and analysis of battery status, when battery problems occur, operation managers cannot accurately judge the remaining battery power and the vehicle's driving capacity. As a result, after the battery alarm sounds, the vehicle often cannot reach the scheduled destination, and may even be forced to stop operating on the way. This situation not only affects the passengers' travel experience, but also brings great inconvenience to operation management.

[0006] In summary, in existing vehicle operations, there is a problem of insufficient accuracy in battery monitoring, which seriously affects the operational efficiency of electric public vehicles. Summary of the invention

[0007] The embodiments of the present application provide a public vehicle operation management method and related equipment based on big data analysis, which are used to solve the problem of insufficient accuracy of battery monitoring and improve the operation efficiency of electric public vehicles.

[0008] To achieve the above objectives, the embodiments of the present application adopt the following technical solutions:

[0009] In a first aspect, a public vehicle operation management method based on big data analysis is provided, which is applied to an electronic device, wherein the electronic device is deployed with a big data model, and the electronic device is connected to a vehicle terminal of an electric public vehicle, and the method comprises:

[0010] In response to receiving a battery alarm signal of the electric public vehicle, obtaining battery parameters and vehicle operation data of the electric public vehicle, wherein the vehicle operation data includes real-time road condition data, vehicle operation data, and driving behavior data of the electric public vehicle on a current driving route; and

[0011] Get current environment data;

[0012] Inputting the current environment data and the real-time road condition data into the big data model to obtain a current vehicle operation scenario, wherein the big data model is constructed based on historical environment data and historical road condition data;

[0013] In a preset battery parameter database, query and obtain a battery parameter range corresponding to the current operation scenario of the vehicle;

[0014] When the battery parameter is within the battery parameter range, the vehicle energy consumption data of the electric public vehicle within a preset kilometer range is obtained, and based on the battery parameter, the vehicle energy consumption data and the current operation scenario of the vehicle, the energy consumption trend of the electric public vehicle within a future preset time period is predicted;

[0015] generating a power output mode of the electric public vehicle on the current driving route according to the energy consumption trend, the vehicle operation data and the driving behavior data, wherein the power output mode is used to control the power distribution of the electric public vehicle;

[0016] generating a braking adjustment instruction according to the power output mode, wherein the braking adjustment instruction is used to instruct the electric public vehicle to execute a braking strategy;

[0017] The braking adjustment instruction is sent to the vehicle terminal so that the vehicle terminal executes the braking strategy.

[0018] In a possible implementation of the first aspect, the step of constructing the big data model includes:

[0019] Using a preset feature extraction algorithm, extracting features from the historical environment data and the historical road condition data to obtain n first features, where n is an integer greater than 3;

[0020] Using a first preset clustering algorithm, cluster the n first features to obtain m clustering clusters, where 0 < m < n and m is an integer;

[0021] Using the feature extraction algorithm, extract features from the m clustering clusters to obtain p second features, where 0 < p < m and p is an integer;

[0022] Using a second preset clustering algorithm, cluster the p second features to obtain q clustering results, where each clustering result includes a clustering center, and 0 < q < p and q is an integer;

[0023] Using a preset distance algorithm, for each clustering center, calculate the average distance between all the second features and the clustering center, and use the clustering center corresponding to the smallest average distance as the target clustering center, and use the clustering result corresponding to the target clustering center as the target clustering result;

[0024] According to a preset mapping function, map the target clustering result to the corresponding operation scenario;

[0025] Package the feature extraction algorithm, the first preset clustering algorithm, the second preset clustering algorithm, the distance algorithm and the mapping function to obtain the big data model.

[0026] In another possible implementation of the first aspect, the method further includes:

[0027] When the battery parameters are not within the battery parameter range, obtain the current battery level of the electric bus through the vehicle terminal, and calculate the maximum driving distance of the electric bus according to the current battery level;

[0028] Obtain the current location of the electric bus;

[0029] According to the current location, determine the remaining driving distance of the electric bus on the current driving route;

[0030] When the remaining driving distance is greater than the maximum driving distance, calculate the straight-line distance between the current location and the end point of the current driving route, and search for an electric bus charging station within the area with the current location as the center and the straight-line distance as the radius;

[0031] When there are multiple electric bus charging stations, use the electric bus charging station with the shortest distance to the current location as the target charging station;

[0032] A new driving route is generated according to the current location, the target charging station and the end point of the current driving route, and the new driving route is sent to the vehicle terminal.

[0033] In another possible implementation of the first aspect, predicting the energy consumption trend of the electric public vehicle within a future preset time period according to the battery parameters, the vehicle energy consumption data, and the current operation scenario of the vehicle includes:

[0034] The battery parameters, the vehicle energy consumption data and the current operation scenario of the vehicle are input into a pre-trained prediction model to obtain the energy consumption trend of the electric public vehicle within a future preset time period.

[0035] In another possible implementation of the first aspect, generating a power output mode of the electric public vehicle on the current driving route according to the energy consumption trend, the vehicle operation data, and the driving behavior data includes:

[0036] According to the energy consumption trend, the power supply corresponding to each time step is calculated, and the power supply corresponding to each time step is fitted to generate a power supply curve of the electric public vehicle;

[0037] Determining the driving state of the electric public vehicle according to the vehicle operation data, wherein the driving state includes an idle state, an acceleration state, a constant speed state, and a deceleration state;

[0038] Acquiring an energy conversion efficiency curve of the electric public vehicle, and determining an optimal output power of the electric public vehicle in the driving state according to the power supply power curve and the energy conversion efficiency curve;

[0039] Calculating the curvature of the power supply curve in real time;

[0040] When the curvature is greater than a preset threshold, a power output mode is determined according to the driving behavior data and the optimal output power of the electric public vehicle in the driving state.

[0041] In another possible implementation of the first aspect, determining the optimal output power of the electric public vehicle in the driving state according to the power supply power curve and the energy conversion efficiency curve includes:

[0042] According to the power supply power curve and the energy conversion efficiency curve, obtaining a power supply power data segment and a corresponding energy conversion efficiency data segment in the driving state;

[0043] For each time point of the power supply data segment and the corresponding energy conversion efficiency data segment, calculating the actual output power at each time point in the driving state, and generating an output power curve according to the actual output power at each time point;

[0044] According to the output power curve and the energy conversion efficiency curve, an optimal efficiency range is determined, and an average value of the output power corresponding to the optimal efficiency range is calculated, and the average value is used as the optimal output power of the electric public vehicle in the driving state.

[0045] In another possible implementation of the first aspect, determining the power output mode according to the driving behavior data and the optimal output power of the electric public vehicle in the driving state includes:

[0046] Determining a driving behavior type according to the driving behavior data, wherein the driving behavior type includes smooth, medium and aggressive;

[0047] Acquire a weight coefficient corresponding to the driving behavior type in a preset database;

[0048] Calculating a power adjustment coefficient of the actual output power corresponding to each time point according to the weight coefficient, wherein the product of the optimal output power and the power adjustment coefficient is less than the maximum output power of the electric public vehicle;

[0049] Calculating the product of the actual output power at each time point and the corresponding power adjustment coefficient to obtain a plurality of adjusted output power values, and generating an optimized power output curve according to all processed output power values;

[0050] Determining a power output mode according to the optimal output power and the optimized power output curve;

[0051] Among them, the power output curve includes a low power interval, a medium power interval and a high power interval, and the power output mode includes an energy-saving mode, a standard mode and a power mode. When the optimal output power is located in the low power interval of the power output curve, the power output mode is the energy-saving mode, when the optimal output power is located in the medium power interval of the power output curve, the power output mode is the standard mode, and when the optimal output power is located in the high power interval of the power output curve, the power output mode is the power mode.

[0052] In another possible implementation manner of the first aspect, the driving behavior data includes a frequency of sudden acceleration and sudden braking, a vehicle turning angle, a vehicle turning speed, and a vehicle idling time, and determining the driving behavior type according to the driving behavior data includes:

[0053] Extracting features of the frequency of sudden acceleration and sudden braking, the vehicle turning angle, the vehicle turning speed, and the vehicle idling time to obtain a plurality of driving behavior features;

[0054] The driving behavior characteristics are input into a pre-trained classification model to obtain the driving behavior type.

[0055] In a second aspect, the present application provides an electronic device, including:

[0056] a memory configured to store instructions; and

[0057] The processor is configured to call the instructions from the memory and implement the above-mentioned public vehicle operation management method based on big data analysis when executing the instructions.

[0058] In a third aspect, the present application provides a public vehicle operation management system based on big data analysis, including:

[0059] Electronic equipment;

[0060] The vehicle terminal of the electric public vehicle is connected to the electronic device.

[0061] Through the above technical solution, accurate monitoring and intelligent management of the battery status are achieved by connecting the electronic device to the vehicle terminal of the electric public vehicle, thereby significantly improving the operational efficiency and safety of the electric public vehicle. After receiving the battery alarm signal, the battery parameters and vehicle operation data of the electric public vehicle are obtained, and the current environmental data is obtained at the same time, and the above data is input into the big data model built based on historical environmental data and historical road condition data to generate the current operation scenario of the vehicle. By querying the battery parameter interval corresponding to the current operation scenario in the preset battery parameter database, it is possible to accurately determine whether the battery status is normal. If the battery parameters are within the normal range, the vehicle energy consumption data within the preset kilometer range will be further obtained, and the energy consumption trend within the preset time period in the future will be predicted by combining the battery parameters, vehicle energy consumption data and the current operation scenario. According to the energy consumption trend, vehicle operation data and driving behavior data, the power output mode of the electric public vehicle on the current driving route is generated to optimize power distribution, and a braking adjustment instruction is generated to instruct the vehicle terminal to execute the braking strategy, thereby realizing intelligent control of the electric public vehicle. It not only improves the accuracy of battery monitoring and reduces the occurrence of false alarms, but also ensures that the vehicle can operate stably under various operating conditions through intelligent prediction and optimization of power distribution, avoiding vehicle shutdowns due to battery problems, and improving passengers' travel experience and convenience of operation management. In addition, through the analysis of real-time road condition data and driving behavior data, the energy consumption management of the vehicle is optimized, the operating cost is further reduced, and strong support is provided for the sustainable development of electric public vehicles.

[0062] Other features and advantages of the embodiments of the present application will be described in detail in the subsequent specific implementation section. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 A flowchart of a public vehicle operation management method based on big data analysis provided in an embodiment of the present application;

[0064] Figure 2 A schematic diagram of the first clustering provided in an embodiment of the present application;

[0065] Figure 3 A schematic diagram of the second clustering provided in an embodiment of the present application;

[0066] Figure 4 A schematic diagram of a flow chart after determining a power output mode provided in an embodiment of the present application;

[0067] Figure 5 A structural block diagram of a public vehicle operation management system based on big data analysis provided in an embodiment of the present application. DETAILED DESCRIPTION

[0068] In order to make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. It should be understood that the specific implementation methods described herein are only used to illustrate and explain the embodiments of the present application, and are not used to limit the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0069] It should be noted that if the embodiments of the present application involve directional indications (such as up, down, left, right, front, back...), the directional indications are only used to explain the relative position relationship, movement status, etc. between the components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indication will also change accordingly.

[0070] In addition, if there are descriptions involving "first", "second", etc. in the embodiments of the present application, the descriptions of "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or suggesting their relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in the field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such combination of technical solutions does not exist and is not within the scope of protection required by this application.

[0071] Embodiment 1, Figure 1 The following schematically shows a flow chart of a method for public vehicle operation management based on big data analysis according to an embodiment of the present application. Figure 1 As shown, an embodiment of the present application provides a public vehicle operation management method based on big data analysis, which is applied to an electronic device, where a big data model is deployed, and the electronic device is connected to a vehicle terminal of an electric public vehicle. The method may include the following steps.

[0072] S110, in response to receiving a battery alarm signal of the electric public vehicle, obtaining battery parameters and vehicle operation data of the electric public vehicle, wherein the vehicle operation data includes real-time road condition data, vehicle operation data, and driving behavior data of the electric public vehicle on a current driving route; and

[0073] S120, obtaining current environment data;

[0074] S130, inputting current environment data and real-time road condition data into a big data model to obtain a current vehicle operation scenario, wherein the big data model is constructed based on historical environment data and historical road condition data;

[0075] S140, querying a preset battery parameter database to obtain a battery parameter range corresponding to the current operation scenario of the vehicle;

[0076] S150, when the battery parameters are within the battery parameter range, obtaining vehicle energy consumption data of the electric public vehicle within a preset kilometer range, and predicting the energy consumption trend of the electric public vehicle within a future preset time period based on the battery parameters, the vehicle energy consumption data and the current operation scenario of the vehicle;

[0077] S160, generating a power output mode of the electric public vehicle on the current driving route according to the energy consumption trend, the vehicle operation data and the driving behavior data, wherein the power output mode is used to control the power distribution of the electric public vehicle;

[0078] S170, generating a braking adjustment instruction according to the power output mode, the braking adjustment instruction being used to instruct the electric public vehicle to execute a braking strategy;

[0079] S180. Send a braking adjustment instruction to the vehicle terminal so that the vehicle terminal executes a braking strategy.

[0080] In this embodiment, battery parameters include temperature, voltage, current, etc.; real-time road condition data include the road type of the current driving route, the traffic density of the current road, etc.; vehicle operation data include vehicle speed, acceleration, motor power output, etc.; driving behavior data include the frequency and intensity of sudden acceleration and braking, turning angle and speed, idling time, etc., where idling time refers to the time when the vehicle is stationary but the engine is running.

[0081] The electronic device first responds to the battery alarm signal received from the electric public vehicle. The battery alarm signal is triggered by the vehicle's battery management system (BMS) when an abnormal condition is detected, such as excessive battery temperature, abnormal voltage, or excessive current. After receiving the alarm signal, the battery parameters, including temperature, voltage, and current, are first collected. The battery parameters can be collected in real time by on-board sensors and transmitted to the electronic device through the vehicle communication network. Secondly, the vehicle operation data is collected. The vehicle operation data includes three main aspects: real-time road condition data, vehicle operation data, and driving behavior data. Real-time road condition data covers the road type of the current driving route (such as highways, urban roads, or mountain roads) and the traffic density of the current road. Real-time road condition data can be obtained through the on-board GPS and real-time traffic information services. Vehicle operation data includes vehicle speed, acceleration, and motor power output. Vehicle operation data can be read directly from the vehicle's control system. Driving behavior data includes the frequency and intensity of sudden acceleration and braking, the angle and speed of turning, and idling time. Driving behavior data can be obtained through the vehicle terminal.

[0082] After receiving the battery alarm signal, the current environmental data is also obtained. Environmental data refers to external factors that affect vehicle performance and battery efficiency. The current environmental data includes but is not limited to the following aspects: First, temperature data, including ambient temperature and road surface temperature. The ambient temperature can be directly measured by the on-board temperature sensor, while the road surface temperature is obtained through the meteorological service. Temperature data has a direct impact on battery performance because the chemical reaction and efficiency of the battery will change with temperature. Secondly, humidity data can be measured by the on-board humidity sensor. Humidity affects air density, which in turn affects the aerodynamic performance and energy consumption of the vehicle. Thirdly, wind speed and wind direction data can be obtained through the on-board anemometer or from the local weather station. Wind speed and wind direction will directly affect the driving resistance of the vehicle, thereby affecting energy consumption. In addition, environmental data also includes altitude, which can be measured by GPS or barometer. Altitude affects air density, which in turn affects engine efficiency and energy consumption.

[0083] The current environmental data and real-time road condition data obtained above are input into a pre-established big data model to determine the current operating scenario of the vehicle. In this embodiment, the big data model is constructed based on a large amount of historical environmental data and historical road condition data. The big data model can use a deep learning algorithm to identify and classify different operating scenarios. Specifically, the input of the big data model includes current environmental data (such as temperature, humidity, wind speed, altitude, etc.) and real-time road condition data (such as road type, traffic density, etc.). The big data model first pre-processes the input data, including data cleaning, standardization, and feature extraction. Then, the pre-processed data can be input into a multi-layer neural network. The multi-layer neural network analyzes the input data, compares and matches the current situation with the historical data, and finally outputs a most matching operating scenario category. For example, the big data model can identify the current situation as a "high temperature and high humidity urban congestion scene" or a "cold and low-speed driving scene".

[0084] After determining the current operating scenario of the vehicle, query the battery parameter range corresponding to the scenario in the preset battery parameter database. The battery parameter database records the normal operating parameter range of the battery under different operating scenarios. During the query process, first use the current operating scenario as a keyword to locate the corresponding record in the database. The record includes the normal variation range of battery temperature, voltage and current under similar scenarios. For example, for the "high temperature city congestion scenario", the battery parameter range is as follows: the battery temperature should be between 35°C and 45°C, the voltage should be between 350V and 380V, and the current should be between -200A and 200A (negative values ​​indicate charging, positive values ​​indicate discharging).

[0085] It should be noted that the battery parameter range is not fixed and can change according to specific scenarios. For example, the battery parameter range can be dynamically adjusted according to ambient temperature, driving speed, road slope, etc. By querying the battery parameter range, you can determine whether the current working status of the battery is within the normal range. If the actual parameters fall within the range, it means that the battery is working normally in the current scenario; if it exceeds the range, further analysis and corresponding measures are required.

[0086] After confirming that the battery parameters are in the normal range, it indicates that the battery may have a false alarm phenomenon. In order to more accurately judge the status of the battery at this time, the battery energy consumption can be predicted to determine the battery status to improve operational efficiency. First, obtain the vehicle energy consumption data within the preset kilometer range. This preset kilometer range is generally the most recent driving distance, such as the last 10 kilometers or 20 kilometers. The specific length can be set according to actual needs. The energy consumption data includes the average energy consumption rate (kWh / km) within the preset kilometer range, the change in the instantaneous energy consumption rate, etc. Then, taking into account the current battery parameters, the acquired energy consumption data, and the current operating scenario of the vehicle, a prediction model is used to estimate the energy consumption trend in the future preset time period (such as the next 1 hour or 2 hours).

[0087] Specifically, the prediction model can be a neural network model, and the prediction results include two main indicators: battery consumption rate and remaining mileage. The battery consumption rate is expressed as the power loss rate per hour or per kilometer, while the remaining mileage is the estimated drivable distance calculated based on the current power and the predicted consumption rate. It should be noted that the prediction is dynamically updated, and the actual energy consumption can be continuously monitored and compared with the prediction results. The prediction model can be continuously adjusted and optimized according to the deviation to improve the accuracy of the prediction.

[0088] Based on the predicted energy consumption trend, combined with the vehicle operation data and driving behavior data, the power output mode of the electric public vehicle on the current driving route is generated. The power output mode is a dynamic strategy used to optimize the power distribution of the vehicle to achieve the goal of maximizing energy efficiency and optimizing battery life. Among them, the energy consumption trend includes the battery consumption rate and the remaining mileage. If the prediction shows that the remaining mileage is low or the battery consumption rate is fast, a more energy-saving power output mode is generated. Secondly, the vehicle operation data can be analyzed, such as the current speed, acceleration, and motor power output. For example, when cruising at high speed, a stable power output mode is generated; while in urban sections with frequent acceleration and deceleration, the power distribution during the acceleration process can be optimized to reduce energy loss. In addition, if frequent sudden acceleration and braking behaviors are detected, a smoother power output curve can be generated to suppress uneconomical driving behaviors.

[0089] Specifically, the power output mode can be a time series of power values, such as maximum power limit, acceleration slope limit, etc. The power output mode can be dynamically adjusted according to the changes in road conditions and real-time feedback of driving behavior. For example, higher power output can be temporarily allowed on uphill sections, and the proportion of energy recovery can be increased on downhill sections. Through this intelligent power output management, it is possible to maximize energy utilization efficiency, extend battery life, and improve the overall operational efficiency of the vehicle while ensuring vehicle performance.

[0090] After the power output mode is generated, a specific braking adjustment instruction can be generated according to the power output mode. The braking adjustment instruction is used to guide the electric public vehicle to implement a specific braking strategy to achieve optimal energy management and driving safety. Specifically, the ideal deceleration curve can be determined according to the power output mode first. The deceleration curve needs to consider not only energy efficiency but also passenger comfort. Then, the optimal braking force distribution can be calculated based on the current vehicle speed and the expected deceleration. The optimal braking force distribution includes the force distribution between conventional braking (such as friction braking) and regenerative braking. At low speeds, regenerative braking can be relied on more to maximize energy recovery; at high speeds or in emergencies, conventional braking can be used more to ensure safety.

[0091] In addition, the braking adjustment command also needs to take road conditions into consideration. For example, on downhill sections, the proportion of regenerative braking can be increased to better control the vehicle speed and recover kinetic energy. On slippery roads, gentler braking commands can be generated to avoid tire slippage. The final generated braking adjustment command is a data packet containing multiple parameters, which may include information such as braking force size, braking force distribution ratio, braking time curve, etc. The braking adjustment command is designed to be directly executed by the vehicle's control terminal to ensure precise control and smooth implementation of the braking process.

[0092] Finally, the generated brake adjustment instructions are sent to the vehicle terminal so that the vehicle terminal executes the corresponding braking strategy. First, the brake adjustment instructions are transmitted from the electronic device to the vehicle terminal (ECU) of the vehicle through the vehicle's communication network (such as CAN bus or Ethernet). After receiving the instructions, the ECU will quickly verify and process them to ensure the effectiveness and safety of the instructions. Then, the ECU will convert these instructions into specific control signals and send them to various executions of the vehicle, including braking, power and energy recovery. In other words, when the vehicle terminal receives the brake adjustment instructions, the vehicle's control terminal first parses the brake adjustment instructions and converts them into specific execution commands. The above execution commands will be distributed to various related sub-units of the vehicle, such as braking, power and battery management.

[0093] In this embodiment, by connecting an electronic device to the vehicle terminal of an electric public vehicle, precise monitoring and intelligent management of the battery state are achieved, thereby significantly improving the operation efficiency and safety of the electric public vehicle. After receiving a battery alarm signal, battery parameters and vehicle operation data of the electric public vehicle are obtained. Meanwhile, current environmental data is obtained, and the above data is input into a big data model constructed based on historical environmental data and historical road condition data to generate the current operation scenario of the vehicle. By querying the battery parameter range corresponding to the current operation scenario in a preset battery parameter database, it is possible to accurately determine whether the battery state is normal. If the battery parameters are within the normal range, vehicle energy consumption data within a preset kilometer range is further obtained, and combined with the battery parameters, vehicle energy consumption data, and the current operation scenario, the energy consumption trend in a future preset time period is predicted. According to the energy consumption trend, vehicle operation data, and driving behavior data, a power output mode of the electric public vehicle on the current driving route is generated for optimizing power distribution, and a braking adjustment instruction is generated to instruct the vehicle terminal to execute a braking strategy, thereby realizing intelligent control of the electric public vehicle. This not only improves the accuracy of battery monitoring and reduces the occurrence of false alarms, but also ensures stable operation of the vehicle under various operating conditions through intelligent prediction and optimized power distribution, avoiding vehicle outages caused by battery problems and enhancing the travel experience of passengers and the convenience of operation management. In addition, through the analysis of real-time road condition data and driving behavior data, the energy consumption management of the vehicle is optimized, further reducing the operation cost and providing strong support for the sustainable development of electric public vehicles.

[0094] In one implementation manner of this embodiment, the steps of constructing the big data model include:

[0095] S210. Using a preset feature extraction algorithm, extract features from historical environmental data and historical road condition data to obtain n first features, where n is an integer greater than 3;

[0096] S220. Using a first preset clustering algorithm, cluster the n first features to obtain m clustering clusters, where 0 < m < n and m is an integer;

[0097] S230. Using a feature extraction algorithm, extract features from the m clustering clusters to obtain p second features, where 0 < p < m and p is an integer;

[0098] S240. Using a second preset clustering algorithm, cluster the p second features to obtain q clustering results, where each clustering result includes a clustering center, and 0 < q < p and q is an integer;

[0099] S250, using a preset distance algorithm, for each cluster center, calculating the average distance between all second features and the cluster center, and taking the cluster center corresponding to the smallest average distance as the target cluster center, and taking the clustering result corresponding to the target cluster center as the target clustering result;

[0100] S260, mapping the target clustering result to the corresponding operation scenario according to a preset mapping function;

[0101] S270, encapsulating the feature extraction algorithm, the first preset clustering algorithm, the second preset clustering algorithm, the distance algorithm and the mapping function to obtain a big data model.

[0102] The feature extraction algorithm is used to extract the most representative and distinguishing features from the original, high-dimensional data to reduce the complexity of the data while retaining key information. In this embodiment, the principal component analysis (PCA) algorithm or the independent component analysis (ICA) algorithm can be used to extract features from the historical environmental data and the historical road condition data, and finally output n first features, where n is an integer greater than 3. For example, n is 15, including 5 environment-related features (such as temperature trends, humidity fluctuations) and 10 road condition-related features (such as traffic flow cycles, road slope distribution).

[0103] The n first features are clustered using a first preset clustering algorithm to group data points with similar features, thereby discovering the intrinsic structure and pattern in the data. The first preset clustering algorithm may be K-means, etc. Taking K-means as an example, it can assign data points to the nearest cluster center through an iterative process, and continuously update the cluster center until convergence.

[0104] Figure 2 FIG. 1 shows a schematic diagram of the first clustering provided by an embodiment of the present application. The output of the first clustering is a plurality of initial clusters based on the original data features. Each cluster represents a set of data points with similar features. Figure 2As shown, in actual implementation, first, it is necessary to determine the appropriate number of clusters m according to the preset algorithm. Assume that the optimal number of clusters is determined to be 10. Then, K-means clustering is performed on the n first features, and k=10 is set. The clustering process divides all data points in the n feature space into 10 clusters. Each cluster represents a combination of data with similar features. For example, the clusters may be: 1. Smooth highway (high speed, low traffic volume, good weather); 2. Congested highway (low speed, high traffic volume, good weather); 3. Normal medium speed (medium speed and traffic volume, good weather); 4. Low speed in rainy days (low speed, medium traffic volume, precipitation); 5. Low speed in heavy fog (low speed, low traffic volume, low visibility); 6. Smooth traffic at night (high speed, low traffic volume, night); 7. Congested during morning rush hour (low speed, high traffic volume, morning period); 8. Congested during evening rush hour (low speed, high traffic volume, evening period); 9. Congested shopping area on weekends (low speed, high traffic volume, weekends, commercial area); 10. Smooth traffic in suburbs on holidays (high speed, medium traffic volume, holidays, suburbs).

[0105] The final output is m clusters, where m is less than n and is a positive integer. Each cluster contains a group of data points with similar characteristics.

[0106] After obtaining m clusters, the feature extraction algorithm is used again to extract features from these clusters, so as to further refine and abstract the data features, so as to extract higher-level and more representative features from the clustering results, so as to facilitate subsequent analysis and model building.

[0107] Feature extraction algorithms can use statistical methods (such as mean, variance, skewness, kurtosis, etc.) to describe the distribution characteristics of each cluster.

[0108] Taking statistical methods as an example, for each cluster, the following features can be calculated: 1. Cluster center coordinates: reflecting the average position of the cluster; 2. Cluster radius: reflecting the compactness of the cluster; 3. Variance of data points within the cluster: reflecting the degree of dispersion of data within the cluster; 4. Distance between clusters: reflecting the degree of separation between clusters.

[0109] Assuming there are 10 clusters (m=10), and 5 features are extracted for each cluster, a total of 50 features will be obtained. However, there may be redundant or highly correlated features among these features. Therefore, feature selection is required to retain only the most representative and discriminative features.

[0110] By calculating the correlation between the features, the p most important second features can be selected. For example, 20 features (p=20) are finally output, and the features include the center coordinates, main directions, density, etc. of each cluster. In this embodiment, p second features are output, where p is less than m and is a positive integer. The second feature represents a higher level of abstraction of the original data, which is used to capture the key information in the clustering results and reduce the dimension of the data.

[0111] In order to discover the structure and pattern of the data at a higher level of abstraction, a second preset clustering algorithm is used to cluster the p second features, so as to group similar second features together to form more general categories, thereby revealing potential operational scenarios or patterns in the data. The second clustering is performed on the basis of the first clustering result, and its output is a higher-level clustering result. In this embodiment, the second preset clustering algorithm can be hierarchical clustering.

[0112] Specifically, taking hierarchical clustering as an example, a tree-like clustering structure can be created, allowing clustering results to be observed at different levels. The specific steps are as follows: 1. First, each second feature is regarded as a separate cluster. 2. Calculate the distance between all cluster pairs (Euclidean distance, Manhattan distance, etc. can be used). 3. Find the two clusters with the closest distance and merge them into a new cluster. 4. Repeat steps 2 and 3 until the preset number of clusters is reached or a certain stopping condition is met.

[0113] Assuming there are 20 second features (p=20), through the second preset clustering algorithm, 4 high-level clusters (q=4) can be obtained. The 4 categories represent 4 different operating scenarios or modes, such as Figure 3 shown.

[0114] exist Figure 3 In the example, each clustering result contains a cluster center. For example, the four clustering results represent: 1. Smooth traffic (including the first clusters 1, 6, and 10), characterized by high speed and low to medium traffic volume; 2. General congestion (including the first clusters 2, 3, 7, 8, and 9), characterized by low to medium speed and medium to high traffic volume; 3. Weather influence (including the first clusters 4 and 5), characterized by low speed and significantly affected by weather conditions; 4. Special time period (including the first clusters 6, 7, 8, 9, and 10), characterized by significant influence by time and location factors. The first clustering identifies specific traffic conditions, while the second clustering classifies these conditions into broader scenario types.

[0115] Finally, q clustering results are output, each of which contains a cluster center, where q is less than p and is a positive integer. Finally, the feature set is summarized into several main categories, each of which represents a typical operation scenario or mode.

[0116] Afterwards, the preset distance algorithm is used to calculate the average distance between all second features and each cluster center, and the cluster center with the smallest average distance is selected as the target cluster center, so as to select the most representative clustering result from the q clustering results obtained as the basis for subsequent analysis and application.

[0117] The distance algorithm can be Euclidean distance, cosine similarity, etc. Taking Euclidean distance as an example, the specific steps are as follows: 1. For each cluster center, calculate the Euclidean distances from all p second features to the center. 2. Calculate the average of these distances to obtain the average distance of the cluster center. 3. Repeat steps 1 and 2 to calculate the average distance of all q cluster centers. 4. Compare the q average distances and take the smallest one as the target cluster center.

[0118] The target clustering results represent the most typical and representative patterns or scenarios in the data, reflecting the performance characteristics of electric public vehicles under the most common operating conditions.

[0119] The target clustering result is mapped to the corresponding operation scenario according to the preset mapping function. Specifically, the preset mapping function establishes the correspondence between the clustering result and the specific operation scenario. The mapping function can be a lookup table or a decision tree. For example, if the target clustering result corresponds to "general congestion" (including 2, 3, 7, 8, 9 of the first clustering), then the preset mapping function can be used to map it to the following operation scenarios: regular urban congestion scenarios (characteristics: low to medium speeds, medium to high traffic volume; time: morning and evening rush hours; location: major urban roads, surrounding commercial areas, and main roads connecting residential areas and work areas).

[0120] Finally, the feature extraction algorithm, the first preset clustering algorithm, the second preset clustering algorithm, the distance algorithm and the mapping function are encapsulated to build a complete big data model. In the specific implementation, a main control process can be written, which calls each component in a specific order. For example, feature extraction is performed first, followed by the first clustering, the second clustering, the distance calculation, and finally the mapping function. Each component is designed as an independent module with clearly defined input and output interfaces, so that individual components can be optimized or replaced without affecting the overall. Finally, a complete, end-to-end big data model is obtained.

[0121] This implementation adopts two rounds of clustering process (first preset clustering algorithm and second preset clustering algorithm) to group data, so as to better capture the hierarchical structure and internal relationship of data, more accurately identify and classify different operation scenarios, and map the optimized clustering results to specific operation scenarios, thus improving the interpretability and practicality of model output.

[0122] In one implementation of this embodiment, the following steps are also included:

[0123] S310, when the battery parameter is not within the battery parameter range, obtaining the current power of the electric public vehicle through the vehicle terminal, and calculating the maximum driving distance of the electric public vehicle according to the current power;

[0124] S320, obtaining the current location of the electric public vehicle;

[0125] S330, determining the remaining driving distance of the electric public vehicle on the current driving route according to the current positioning;

[0126] S340, when the remaining driving distance is greater than the maximum driving distance, calculating the straight-line distance between the current location and the end point of the current driving route, and searching for an electric public vehicle charging station in an area with the current location as the center and the straight-line distance as the radius;

[0127] S350: When there are multiple electric public vehicle charging stations, the electric public vehicle charging station with the shortest distance to the current location is used as the target charging station;

[0128] S360: Generate a new driving route based on the current location, the target charging station and the end point of the current driving route, and send the new driving route to the vehicle terminal.

[0129] In this embodiment, it is first determined whether the battery parameters are within the preset parameter range. If not within the range, the current power of the electric public vehicle is obtained through the vehicle terminal, and the maximum driving distance is calculated. In specific implementation, the on-board terminal will read the real-time power data provided by the battery management system (BMS) through the CAN bus or other on-board networks. The real-time power data is usually expressed as a percentage. Subsequently, the power is converted into an estimated mileage based on the energy consumption model specific to the vehicle model. For example, assuming that an electric bus has a full-charge range of 300 kilometers and the current power is 60%, taking into account the influence coefficient of road conditions, weather and other factors of 0.9, the maximum driving distance can be estimated as: 300×60%×0.9≈162 kilometers.

[0130] After that, the current location of the electric public vehicle can be obtained through the on-board GPS module or other positioning systems, such as the Beidou navigation system. In specific implementation, the on-board terminal will regularly (usually every second or every few seconds) update the location information, including data such as longitude, latitude, altitude, speed and direction. These data are transmitted to the central management system (electronic device) in real time through a mobile communication network (such as 4G / 5G). For example, the current location can be: longitude 116.3972°E, latitude 39.9075°N, altitude 50 meters, speed 40km / h, direction 270° (due west).

[0131] Based on the current location obtained, the remaining distance of the electric public vehicle on the current route is determined. In specific implementation, the GPS coordinates of the vehicle need to be mapped to the nearest road section on the digital map. Then, the path planning algorithm is used to calculate the shortest path from the current location to the destination. For example, assuming that the total length of the current route is 50 kilometers, and the current location of the vehicle is determined to be 35 kilometers from the starting point through map matching, the remaining distance is 15 kilometers.

[0132] When the remaining driving distance is greater than the maximum driving distance, the straight-line distance between the current location and the end point of the current driving route is calculated, and the electric public vehicle charging station is searched within this range. Specifically, first, the straight-line distance between the two points is calculated using the Haversine formula. Assuming that the current location is (l at1, lon1) and the end point is (l at2, lon2), the straight-line distance D can be calculated using the following formula:

[0133]

[0134] Where R is the radius of the earth (about 6371 kilometers). For example, if the current location is (39.9075°N, 116.3972°E) and the end point is (39.9868°N, 116.4321°E), the calculated straight-line distance is about 9.38 kilometers.

[0135] Next, with the current location as the center of the circle and the calculated straight-line distance as the radius, a spatial query is performed in the GIS database to filter out all charging stations located in the circular area.

[0136] In the case of finding multiple electric public vehicle charging stations, the charging station with the shortest distance to the current location is selected as the target charging station. In this embodiment, the K nearest neighbor (KNN) algorithm can be used, where K = 1. In specific implementation, the distance from the current location to each candidate charging station is first calculated. The Euclidean distance can be used. Taking the Euclidean distance as an example, for the current location (x0, y0) and the charging station location (x i ,y i ), distance d i The calculation is as follows:

[0137]

[0138] After calculating all distances, the charging station with the shortest distance is selected as the target. For example, if there are three candidate charging stations, and their distances to the current location are 5.2 km, 3.8 km, and 4.5 km, respectively, the charging station with a distance of 3.8 km is selected as the target charging station.

[0139] According to the current location, the target charging station and the end point of the current driving route, a new driving route is generated and sent to the vehicle terminal. Specifically, the interaction between the multi-point path planning algorithm and the vehicle navigation system can be adopted. In the specific implementation, the preset path planning algorithm is first used to calculate the optimal route through the charging station. The algorithm takes into account factors such as the topological structure of the road network, traffic rules, and real-time road conditions. For example, assuming that the current position is A, the target charging station is B, and the end point is C, the algorithm will calculate the optimal path from A to B, and then from B to C, and merge the two segments into a complete new route. The new route may be represented as a sequence of coordinates of a series of waypoints or a sequence of road segment IDs. The generated route data usually includes information such as total distance, estimated time, and turn instructions. Subsequently, this data is sent to the vehicle terminal via a wireless communication network (such as 4G / 5G). After the vehicle terminal receives the new route, it updates the navigation information and informs the driver of the route change through voice prompts.

[0140] This implementation method can intelligently determine the charging demand and automatically plan the optimal charging route by real-time monitoring of the power status, location information and driving route of electric public vehicles, effectively solving the problem of electric vehicle range limitation and improving the flexibility and reliability of vehicle scheduling. Through precise route planning and charging station selection, the operating efficiency of the vehicle is maximized, while the risk of service interruption due to power exhaustion is reduced, effectively improving the operating efficiency of electric public vehicles.

[0141] In one implementation of this embodiment, based on battery parameters, vehicle energy consumption data and the current vehicle operation scenario, predicting the energy consumption trend of an electric public vehicle within a preset time period in the future includes the following steps:

[0142] S410: Input the battery parameters, vehicle energy consumption data and the current operation scenario of the vehicle into a pre-trained prediction model to obtain the energy consumption trend of the electric public vehicle in a future preset time period.

[0143] In this embodiment, the battery parameters, vehicle energy consumption data and the current operation scenario of the vehicle are input into the pre-trained prediction model to obtain the energy consumption trend of the electric public vehicle in the future preset time period. In specific implementation, the battery parameters include key indicators such as battery capacity, current power, number of charge and discharge, and battery internal resistance. The vehicle energy consumption data includes historical average energy consumption, air conditioning power consumption, climbing energy consumption, etc. The current operation scenario of the vehicle includes environmental factors that affect energy consumption, such as current road conditions, weather conditions, and driving speed.

[0144] The prediction model can use a deep learning model, such as a long short-term memory network (LSTM), to process time series data and capture long-term dependencies. The trained deep learning model can predict the energy consumption trend for a period of time in the future (such as the next 2 hours or 4 hours) based on the current state of the input.

[0145] In actual applications, assuming that the LSTM model is used, the input vector contains 20 features: 5 battery parameters (such as battery capacity, current power, battery temperature, internal resistance, and number of cycles), 5 vehicle energy consumption data (such as average energy consumption in the past hour, air conditioning power, motor efficiency, vehicle speed, and acceleration), and 10 operating scenario parameters (such as ambient temperature, humidity, wind speed, road congestion index, slope, passenger load, etc.). The model output is the estimated energy consumption every 15 minutes in the next 4 hours, a total of 16 values.

[0146] For example, the current status of an electric bus is as follows: battery parameters: 100kWh capacity, current power 80%, battery temperature 25℃, internal resistance 0.1Ω, cycle number 500 times; energy consumption data: average energy consumption in the past hour 1.2kWh / km, air conditioning power 3kW, motor efficiency 90%, current speed 40km / h, acceleration 0.5m / s 2 ; Operation scenario: ambient temperature 30℃, humidity 60%, wind speed 2m / s, traffic congestion index 0.3, average slope 1%, passenger load factor 70%, etc. The above data is standardized and input into the pre-trained LSTM model, and the following prediction results are output: Estimated energy consumption every 15 minutes in the next 4 hours (kWh): [1.8, 1.7, 1.9, 2.0, 2.1, 2.0, 1.8, 1.7, 1.6, 1.5, 1.4, 1.3, 1.2, 1.1, 1.0, 0.9].

[0147] The result shows that energy consumption may increase slightly in the next hour (possibly due to entering a congested section or an uphill section), and then gradually decrease (possibly due to entering a flat section).

[0148] By accurately predicting future energy consumption trends, this implementation method can plan charging strategies in advance to avoid service interruptions caused by power exhaustion, thereby greatly improving the operational efficiency and reliability of electric public vehicles.

[0149] In one implementation of this embodiment, generating a power output mode of an electric public vehicle on a current driving route according to energy consumption trends, vehicle operation data, and driving behavior data includes the following steps:

[0150] S510, calculating the power supply corresponding to each time step according to the energy consumption trend, and fitting the power supply corresponding to each time step to generate a power supply curve of the electric public vehicle;

[0151] S520, determining the driving state of the electric public vehicle according to the vehicle operation data, wherein the driving state includes an idling state, an accelerating state, a constant speed state, and a decelerating state;

[0152] S530, obtaining an energy conversion efficiency curve of the electric public vehicle, and determining the optimal output power of the electric public vehicle in a driving state according to the power supply power curve and the energy conversion efficiency curve;

[0153] S540, calculating the curvature of the power supply curve in real time;

[0154] S550: When the curvature is greater than a preset threshold, determine a power output mode according to the driving behavior data and the optimal output power of the electric public vehicle in the driving state.

[0155] First, the power supply for each time step is calculated based on the energy consumption trend, and the power supply corresponding to each time step is fitted to generate the power supply curve of the electric public vehicle. Specifically, the required power supply is calculated for each time step (for example, every 15 minutes) based on the energy consumption trend. The calculation method is to divide the energy consumption of each time interval by the time interval to obtain the average power. For example, if 3kWh of energy is expected to be consumed in 15 minutes, the average power supply for this time step is 12kW (3kWh / 0.25h).

[0156] Next, curve fitting is performed on the discrete power points. Fitting methods include polynomial fitting, spline interpolation or Fourier series fitting. In this embodiment, cubic spline interpolation is used to ensure the smoothness and continuity of the curve. Assume that there are the following power data points: (0, 10kW), (15, 12kW), (30, 15kW), (45, 13kW), (60, 11kW), where the first number represents time (minutes) and the second number represents power. After using cubic spline interpolation, a continuous function P(t) can be obtained, which represents the power supply power at any time t.

[0157] The driving state of the electric public vehicle is determined based on the vehicle operation data, including the idle state, acceleration state, constant speed state and deceleration state. First, the vehicle speed and acceleration data are acquired in real time through on-board sensors (such as speed sensors, accelerometers). The rules for determining the driving state of the electric public vehicle are as follows: 1. Idle state: the vehicle speed is close to 0 (for example, less than 1km / h), and the acceleration is close to 0. 2. Acceleration state: the acceleration is greater than a certain positive threshold (such as 0.1m / s 2 ). 3. Constant speed state: The vehicle speed is greater than a certain threshold (such as 5km / h), and the acceleration is close to 0 (for example, the absolute value is less than 0.05m / s 2 ). 4. Deceleration state: acceleration is less than a negative threshold (such as -0.1m / s 2 ).

[0158] Afterwards, the energy conversion efficiency curve of the electric public vehicle is obtained, and the optimal output power of the electric public vehicle in each driving state is determined according to the power supply power curve and the energy conversion efficiency curve.

[0159] First, the energy conversion efficiency curve of the motor is obtained through a preset energy conversion efficiency curve database. The energy conversion efficiency curve is expressed as the relationship between the motor output power and efficiency. For example, a typical motor efficiency curve may have the highest efficiency (such as 90-95%) in the range of 20-80% rated power, and lower efficiency at extremely low or extremely high power.

[0160] Next, the power supply curve is combined with the energy conversion efficiency curve to determine the optimal output power for each driving state. The specific implementation method can use a table lookup method. For example, assuming that at a certain moment, the power given by the power supply curve is 50kW, and the efficiency curve is found to be the highest efficiency (reaching 94%) at 48kW. Then, 48kW is used as the optimal output power at that moment.

[0161] For different driving conditions, the selection criteria for optimal power may be different: 1. Idle state: Select the lowest power that can maintain basic system operation. 2. Acceleration state: Select the power point with the highest efficiency while meeting the acceleration requirements. 3. Constant speed state: Select the power point that can maintain the current speed and has the highest efficiency. 4. Deceleration state: Give priority to energy recovery and select the power point that maximizes recovery efficiency.

[0162] The curvature of the power supply curve is calculated in real time. Curvature is an important indicator that describes the degree of curvature of the curve, reflecting the degree to which the curve deviates from a straight line near a certain point.

[0163] For a parameterized plane curve, such as the power supply curve P(t), where t represents time, the curvature K can be calculated by the following formula:

[0164] K=|P″(t)| / (1+(P′(t)) 2 ) (3 / 2) ;

[0165] where P'(t) and P"(t) are the first and second order derivatives of P(t), respectively.

[0166] In practical applications, since the power supply curve is usually a discrete data point, it is necessary to use numerical methods to approximate the derivative and curvature. The central difference method can be used:

[0167] P′(t)≈(P(t+h)-P(th)) / (2h);

[0168] P″(t)≈(P(t+h)-2P(t)+P(th) / h2 ;

[0169] Where h is the time step. For example, suppose at time t = 30 minutes, P(29) = 48 kW, P(30) = 50 kW, P(31) = 53 kW, and the time step h = 1 minute. Then:

[0170] P'(30)≈(53-48) / 2=2.5kW / min;

[0171] P”(30)≈(53-2×50+48)=1kW / min 2 ;

[0172] Substituting into the curvature formula, the curvature at t = 30 minutes can be calculated.

[0173] The curvature can be used to monitor the severity of power supply changes in real time. High curvature means rapid changes in power demand. By monitoring the curvature, sudden changes in power demand can be predicted, thereby improving the flexibility and response speed of energy management.

[0174] When the curvature is greater than a preset threshold, the power output mode is determined based on the driving behavior data and the optimal output power of the electric public vehicle in the current driving state. First, a curvature threshold is set, which can be set to twice the average curvature. When the curvature is detected to exceed the threshold, the power management strategy is triggered.

[0175] Next, the driving behavior data is analyzed. The driving behavior data includes the driver's accelerator pedal position, braking frequency, steering angle, etc. For example, if the driver frequently steps on the accelerator pedal, it is predicted that the electric public vehicle is about to enter a section of road that requires frequent acceleration.

[0176] The power output mode is determined by combining driving behavior data and the optimal output power determined previously. The power output modes may include: 1. Energy-saving mode: Limiting power output to a lower level to maximize energy efficiency while ensuring basic performance. 2. Dynamic response mode: Allowing rapid power changes to accommodate frequent acceleration and deceleration needs. 3. Smooth mode: Smoothing the energy consumption curve through gradual changes in power output to reduce peak loads. 4. Performance mode: Allowing higher power output for a short period of time to meet instantaneous high performance needs.

[0177] For example, assuming the current curvature is 0.5, which exceeds the preset threshold of 0.3. Driving behavior data shows that the driver is accelerating frequently, and the optimal output power for the current driving state is 60kW, the dynamic response mode is selected to set the power output range to 50-70kW in order to quickly respond to the driver's operation while maintaining fluctuations near the optimal power.

[0178] This implementation method can adjust the power output more flexibly by considering the change trend (curvature) of the power supply, driving behavior and optimal power, which can not only meet driving needs but also optimize energy utilization within the possible range. It can improve the dynamic performance of the vehicle while maximizing energy efficiency, thus achieving a better balance between performance and efficiency and realizing intelligent and personalized management of power output.

[0179] In one implementation of this embodiment, determining the optimal output power of the electric public vehicle in a driving state according to the power supply power curve and the energy conversion efficiency curve includes the following steps:

[0180] S610, acquiring a power supply data segment and a corresponding energy conversion efficiency data segment in a driving state according to a power supply power curve and an energy conversion efficiency curve;

[0181] S620, for each time point of the power supply data segment and the corresponding energy conversion efficiency data segment, calculating the actual output power at each time point in the driving state, and generating an output power curve according to the actual output power at each time point;

[0182] S630: Determine the optimal efficiency range according to the output power curve and the energy conversion efficiency curve, calculate the average value of the output power corresponding to the optimal efficiency range, and use the average value as the optimal output power of the electric public vehicle in the driving state.

[0183] First, according to the power supply power curve and the energy conversion efficiency curve, obtain the power supply power data segment and the corresponding energy conversion efficiency data segment under a specific driving state. Specifically, it is necessary to locate the corresponding time period in the entire power supply power curve according to the driving state (such as idling, acceleration, uniform speed or deceleration). This can be achieved by finding the time point when the driving state changes. For example, if it is detected that the time point when the vehicle changes from the idle state to the accelerated state is t1, and the time point when it changes to the uniform speed state again is t2, then the power supply power data in the time interval [t1, t2] corresponds to the accelerated state.

[0184] Next, extract the data of the time interval from the power supply curve to form the power supply data segment of the driving state. At the same time, find the efficiency data corresponding to these power supply values ​​in the energy conversion efficiency curve. The energy conversion efficiency curve is a function with power as the independent variable, so the corresponding efficiency value can be directly obtained by table lookup or interpolation. For example, assuming that in the acceleration state, the power supply power at time t1 is 40kW and at time t2 is 60kW, then find the efficiency values ​​corresponding to all power points in the range of 40kW to 60kW in the energy conversion efficiency curve.

[0185] Afterwards, for each time point of the power supply data segment and the corresponding energy conversion efficiency data segment, the actual output power at each time point in the driving state is calculated, and the output power curve is generated based on these actual output powers. First, the actual output power is the result of the power supply after energy conversion, and its calculation formula is:

[0186] Actual output power = power supply × energy conversion efficiency;

[0187] For each time point, the above calculation is required. For example, assuming that at a certain time point t, the power supply power is 50kW, and the corresponding energy conversion efficiency is 0.9 (ie 90%), then the actual output power at this time point is: 50kW×0.9=45kW.

[0188] Assume the following data: t1: power supply power 40kW, efficiency 0.88; t2: power supply power 45kW, efficiency 0.90; t3: power supply power 50kW, efficiency 0.92; the actual output power calculated is: t1: 40kW×0.88=35.2kW; t2: 45kW×0.90=40.5kW; t3: 50kW×0.92=46.0kW.

[0189] Based on the above calculation results, a new data series can be generated, that is, the change of actual output power over time. By connecting these data points, an output power curve can be formed.

[0190] According to the output power curve and the energy conversion efficiency curve, the optimal efficiency interval is determined, and the average value of the output power corresponding to the interval is calculated, and the average value is used as the optimal output power of the electric public vehicle in a specific driving state. First, the output power curve and the energy conversion efficiency curve need to be compared and analyzed. Usually, the energy conversion efficiency curve will show one or more high-efficiency areas, and the task is to find the part in these areas that best matches the actual output power.

[0191] The method for determining the optimal efficiency interval can adopt the following steps: 1. Set an efficiency threshold on the energy conversion efficiency curve, such as 90%. 2. Find all power intervals with efficiencies above this threshold. 3. Check the overlap of these high-efficiency intervals with the actual output power curve. 4. Select the interval with the highest overlap as the optimal efficiency interval. For example, suppose the energy conversion efficiency curve shows that the efficiency is more than 90% in the range of 35-50kW, and the actual output power curve falls in the range of 30-55kW most of the time. Then, 35-50kW can be determined as the optimal efficiency interval.

[0192] Next, calculate the average output power within the optimal efficiency range. Assume that within this range, the sampled output power values ​​are as follows: 37kW, 40kW, 43kW, 46kW, 49kW; the average value is calculated as: (37+40+43+46+49) / 5=43kW. Then 43kW is determined to be the optimal output power under this driving state.

[0193] Through the output power curve, this embodiment can intuitively see the actual power output capacity of the vehicle at different times, and obtain an optimal output power value that can meet the actual power demand while ensuring high efficiency. By controlling the actual operation near this power, the energy utilization efficiency can be maximized and unnecessary energy loss can be reduced.

[0194] In one implementation of this embodiment, determining the power output mode according to the driving behavior data and the optimal output power of the electric public vehicle in the driving state includes the following steps:

[0195] S710, determining a driving behavior type according to the driving behavior data, wherein the driving behavior types include smooth, medium, and aggressive;

[0196] S720, obtaining a weight coefficient corresponding to the driving behavior type in a preset database;

[0197] S730, calculating a power adjustment coefficient of the actual output power corresponding to each time point according to the weight coefficient, wherein the product of the optimal output power and the power adjustment coefficient is less than the maximum output power of the electric public vehicle;

[0198] S740, calculating the product of the actual output power at each time point and the corresponding power adjustment coefficient to obtain a plurality of adjusted output power values, and generating an optimized power output curve according to all processed output power values;

[0199] S750, determining a power output mode according to the optimal output power and the optimized power output curve;

[0200] Among them, the power output curve includes a low power interval, a medium power interval and a high power interval, and the power output mode includes an energy-saving mode, a standard mode and a power mode. When the optimal output power is in the low power interval of the power output curve, the power output mode is the energy-saving mode, when the optimal output power is in the medium power interval of the power output curve, the power output mode is the standard mode, and when the optimal output power is in the high power interval of the power output curve, the power output mode is the power mode.

[0201] Determine the driving behavior type based on driving behavior data, including three types: smooth, medium, and aggressive. First, extract key features from the driving behavior data. The key features include the acceleration change rate, braking frequency, steering angle change, etc. For example, indicators such as the average acceleration, maximum acceleration, and acceleration standard deviation over a period of time can be calculated. Then, compare these feature values with predefined thresholds to determine the driving behavior type.

[0202] In specific implementation, the following method can be adopted: Set the acceleration change rate thresholds a1 and a2, where a1 < a2. Calculate the average acceleration change rate a over a period of time (such as 5 minutes). If a < a1, it is determined as smooth driving; if a1 ≤ a < a2, it is determined as medium driving; if a ≥ a2, it is determined as aggressive driving. For example, assume that a1 = 2m / s 3 and a2 = 4m / s 3 and the average acceleration change rate calculated over a certain period of time is 3.5m / s 3 , the number of hard brakes is 2 times, and the number of rapid lane changes is 1 time. Judging from the acceleration change rate, it belongs to medium driving.

[0203] Smooth driving means lower energy consumption and higher energy efficiency, while aggressive driving may lead to increased energy consumption and reduced efficiency.

[0204] After that, obtain the weight coefficient corresponding to the driving behavior type in the preset database. The preset database contains different driving behavior types and their corresponding weight coefficients, which reflect the influence degree of different driving behaviors on the power output demand.

[0205] In specific implementation, the following method can be adopted: Store a mapping table in the database, which maps the three driving behavior types of smooth, medium, and aggressive to different weight coefficient ranges respectively. For example: Smooth driving: weight coefficient range 0.8 - 1.0; Medium driving: weight coefficient range 1.0 - 1.2; Aggressive driving: weight coefficient range 1.2 - 1.5.

[0206] When the driving behavior type is determined, the corresponding weight coefficient range can be queried from the database. To more accurately reflect the continuous change of the driving behavior, interpolation calculation can be performed within each range. For example, if it is determined as medium driving but偏向 smooth, 1.05 can be selected as the weight coefficient; if it is偏向 aggressive, 1.15 can be selected as the weight coefficient.

[0207] The power adjustment coefficient of the actual output power corresponding to each time point is calculated according to the weight coefficient, while ensuring that the product of the optimal output power and the power adjustment coefficient is less than the maximum output power of the electric public vehicle. Specifically, the power adjustment coefficient is used to fine-tune the actual output power according to the driving behavior to better adapt to the driver's driving style. In specific implementation, the following method can be used: 1. First, the weight coefficient obtained in step S720 is used as the basic adjustment coefficient. 2. Based on the maximum output power of the vehicle, the power adjustment coefficient is calculated according to the basic adjustment coefficient.

[0208] The power adjustment factor can be calculated using the following formula:

[0209] Power adjustment coefficient = min (weight coefficient, maximum output power / optimal output power);

[0210] Among them, the min function can ensure that the adjusted power does not exceed the maximum output power of the vehicle. For example, assuming that the maximum output power of an electric public vehicle is 200kW, the optimal output power at a certain moment is 150kW, and the weight coefficient is 1.2 (corresponding to aggressive driving). Then: power adjustment coefficient = min(1.2,200kW / 150kW) = min(1.2,1.33) = 1.2.

[0211] In the above example, even with aggressive driving, the adjusted power (150 kW x 1.2 = 180 kW) will not exceed the maximum output power of 200 kW.

[0212] Calculate the product of the actual output power at each time point and the corresponding power adjustment coefficient to obtain multiple adjusted output power values, and generate an optimized power output curve based on all processed output power values. Specifically, first adjust the actual output power at each time point, and then connect the adjusted data points to form a new power output curve. In the specific implementation, the method is as follows: 1. For each time point, multiply the actual output power by the power adjustment coefficient calculated in step S730. 2. Arrange all adjusted power values ​​in chronological order to form a new data sequence. 3. Use a preset interpolation algorithm to connect discrete data points into a continuous curve. For example, assume the following data: time point 1: actual output power 100kW, power adjustment coefficient 1.1; time point 2: actual output power 120kW, power adjustment coefficient 1.15; time point 3: actual output power 110kW, power adjustment coefficient 1.05.

[0213] Calculate the adjusted output power: time point 1: 100kW*1.1=110kW; time point 2: 120kW*1.15=138kW; time point 3: 110kW*1.05=115.5kW.

[0214] The adjusted power values ​​form a new data sequence: [110kW, 138kW, 115.5kW]. Next, the discrete points can be connected into a smooth curve using methods such as cubic spline interpolation or Bezier curves to obtain the optimized power output curve.

[0215] Finally, the power output mode is determined according to the optimal output power and the optimized power output curve. First, the optimized power output curve is divided into a low power interval, a medium power interval and a high power interval. Then, the corresponding power output mode is determined according to the interval where the optimal output power is located. In the specific implementation, the division threshold of the power interval is first determined. For example, the maximum power of 0-40% can be defined as the low power interval, 40-70% can be defined as the medium power interval, and 70-100% can be defined as the high power interval. Secondly, the optimized power output curve is analyzed to determine the range of each interval. Finally, the interval where the optimal output power is located is determined, and the corresponding power output mode is selected accordingly. For example, assuming that the maximum output power of the electric public vehicle is 200kW, the power interval is divided as follows: low power interval: 0-80kW; medium power interval: 80-140kW; high power interval: 140-200kW. If the calculated optimal output power is 110kW, then the optimal output power falls in the medium power interval. According to the rule, the standard mode should be selected as the power output mode at this time.

[0216] The optimized power output curve of this embodiment can better reflect the driver's driving style and needs than the original actual output power curve, and can effectively ensure that the power output will not exceed the physical limitations of the vehicle, thereby more accurately controlling the power output of the vehicle, which not only meets the driver's control needs, but also optimizes energy utilization efficiency. In addition, according to the current optimal working point of the vehicle and the overall power distribution characteristics, the most suitable power output mode can be selected. The energy-saving mode is suitable for low-load conditions and can maximize energy efficiency; the standard mode balances performance and efficiency and is suitable for daily driving; the power mode is enabled when high-performance output is required to meet special driving needs. Therefore, the working state of the vehicle can be dynamically adjusted according to the actual situation, which can not only meet the performance requirements in different scenarios, but also maximize energy efficiency when possible, which not only improves the adaptability of the vehicle, but also provides a better driving experience for the driver.

[0217] In one implementation of this embodiment, the driving behavior data includes the frequency of sudden acceleration and sudden braking, the vehicle turning angle, the vehicle turning speed, and the vehicle idling time. Determining the driving behavior type according to the driving behavior data includes the following steps:

[0218] S810, extracting features of the frequency of sudden acceleration and sudden braking, the vehicle turning angle, the vehicle turning speed, and the vehicle idling time to obtain a plurality of driving behavior features;

[0219] S820: Input the driving behavior features into a pre-trained classification model to obtain a driving behavior type.

[0220] First, multiple key features are extracted from the original driving behavior data. For the frequency of sudden acceleration and braking, the acceleration threshold (such as forward acceleration greater than 3m / s 2 For rapid acceleration, the negative acceleration is less than -3m / s 2 For emergency braking), count the number of times the threshold is exceeded per unit time (such as per hour). For the vehicle turning angle, the steering wheel angle sensor data can be used to calculate the maximum angle value of each turn. The vehicle turning speed can be obtained through the angular velocity sensor, and the maximum angular velocity during each turn can be recorded. The vehicle idling time can be counted by judging the duration of the engine running but the vehicle speed is zero. After obtaining the key features, the sliding window method can be used to calculate the time series features. For example, the 7-day sliding average of the number of emergency accelerations per hour can be calculated to reflect the long-term trend of driving behavior. Finally, a set of feature vectors that can fully reflect the characteristics of driving behavior are obtained, providing input for subsequent classification models.

[0221] Afterwards, the extracted driving behavior features are input into a pre-trained classification model to obtain the driving behavior type. The pre-trained classification model here can use a machine learning algorithm, such as a deep learning model. In the training phase, a large amount of labeled driving behavior data needs to be prepared, including feature vectors and corresponding behavior type labels (such as aggressive, stable, energy-saving, etc.). The model is trained using the above data so that the model learns the mapping relationship between features and behavior types.

[0222] In the prediction phase, new feature vectors are fed into the model, and the model can predict the type of driving behavior based on the learned rules. For example, if a feature vector shows frequent sudden acceleration, sudden braking, large turning angles and speeds, the model will classify it as "aggressive" driving behavior.

[0223] This embodiment achieves accurate quantification and intelligent identification of driving behavior by extracting and classifying features of driving behavior data.

[0224] Embodiment 2, Figure 4 FIG. 1 shows a flow chart of the process after determining the power output mode according to an embodiment of the present application, referring to FIG. Figure 4 After determining the power output mode according to the driving behavior data and the optimal output power of the electric public vehicle in the driving state, the following steps are also included:

[0225] S910, obtaining the current location and current time point of the electric public vehicle, and determining the remaining mileage of the electric public vehicle on the current driving route according to the current location;

[0226] S920, inputting the remaining mileage into the pre-trained prediction model to obtain the end time point when the electric public vehicle reaches the end point of the current driving route;

[0227] S930, calculating a first energy consumption value of the electric public vehicle from the current time point to the end time point according to the optimized power output curve and power output mode;

[0228] S940, determining a second energy consumption value of the remaining mileage according to the energy consumption trend;

[0229] S950, taking the value of the second energy consumption value minus the first energy consumption value as the energy saving amount of the remaining mileage of the electric public vehicle on the current driving route and displaying it;

[0230] According to the optimized power output curve and power output mode, the first energy consumption value of the electric public vehicle from the current time point to the end time point is calculated, including:

[0231] S931, mapping the remaining mileage to the time dimension of the power output curve, and determining a target curve segment from the current time point to the end time point on the power output curve, and dividing the target curve segment into a low power interval, a medium power interval, and a high power interval;

[0232] S932, calculating a first power average value of a low power interval of the target curve segment, and obtaining a first time period of the low power interval;

[0233] S933, calculating a second power average value of a medium power interval of the target curve segment, and obtaining a second time period of the medium power interval;

[0234] S934, calculating a third power average value of a high power interval of the target curve segment, and obtaining a third time period of the high power interval;

[0235] S935. Query the preset power output mode database to obtain a first energy conversion efficiency corresponding to the energy-saving mode, a second energy conversion efficiency corresponding to the standard mode, and a third energy conversion efficiency corresponding to the power mode;

[0236] S936: taking the product of the first power average value and the first time period as the first energy consumption, taking the product of the second power average value and the second time period as the second energy consumption, and taking the product of the third power average value and the third time period as the third energy consumption;

[0237] S937, calculating a first value of a first energy consumption and a first energy conversion efficiency, calculating a second value of a second energy consumption and a second energy conversion efficiency, and calculating a third value of a third energy consumption and a third energy conversion efficiency;

[0238] S938. The sum of the first value, the second value and the third value is used as the energy saving amount of the remaining mileage of the electric public vehicle on the current driving route, and is displayed.

[0239] First, the current location and time of the electric public vehicle are obtained, and the remaining mileage is calculated. This can be achieved through the global positioning system (GPS) and the on-board computer system. The GPS receiver can accurately locate the latitude and longitude coordinates of the vehicle, while the on-board computer records the current timestamp. After obtaining the above data, the current location is compared with the preset driving route to calculate the remaining driving distance. For example, if the total route length of an electric bus is 20 kilometers, and the current GPS shows that the vehicle has traveled 12 kilometers, then the remaining mileage is 8 kilometers.

[0240] After that, the time it will take for the vehicle to reach the destination is estimated using a pre-trained prediction model. The prediction model can be a machine learning algorithm, such as a long short-term memory network (LSTM), which takes into account a variety of factors that affect the travel time. The input data includes not only the remaining mileage, but also relevant data such as the current time, weather conditions, and traffic flow. The prediction model is trained on a large amount of historical driving data and learns the complex relationship between these factors and the actual travel time. For example, if the remaining mileage is 8 kilometers and the current time is 5 pm on a weekday, the model may predict that it will take 30 minutes to reach the destination, taking into account possible traffic congestion. During the prediction process, the model estimates the travel time based on the input remaining mileage and other relevant factors using the learned patterns.

[0241] According to the optimized power output curve and power output mode, the first energy consumption value of the electric public vehicle from the current time point to the destination time point is calculated. The power output curve describes the expected power output at each time point from the current time to the estimated arrival time at the destination.

[0242] Afterwards, the second energy consumption value of the remaining mileage is determined according to the energy consumption trend. The second energy consumption value of the remaining mileage can be obtained by adding up all energy consumptions between the current time point and the end time point.

[0243] The energy saving of the remaining mileage of the electric public vehicle on the current route is obtained by subtracting the first energy consumption value calculated based on the optimized power output from the second energy consumption value predicted based on the energy consumption trend. For example, if the second energy consumption value is 6.4kWh, and the first energy consumption value of the expected energy consumption after optimization is 5.8kWh, then the energy saving is 6.4kWh-5.8kWh=0.6kWh. The energy saving reflects the energy benefit brought by the optimization strategy. After the calculation is completed, the energy saving is displayed in the form of a numerical value. The numerical value can be displayed directly (such as "estimated energy saving of 0.6kWh"), or it can be converted into a percentage ("estimated energy saving of 9.4%) or equivalent mileage ("equivalent to an increase of 1.5 kilometers of driving mileage").

[0244] Specifically, according to the optimized power output curve and power output mode, calculating the first energy consumption value of the electric public vehicle from the current time point to the end time point includes the following steps:

[0245] First, map the remaining mileage to the time dimension of the power output curve and divide the power intervals. First, convert the spatial dimension (remaining mileage) into the time dimension. For example, if the remaining mileage is 10 kilometers and the expected average speed is 30 kilometers per hour, the corresponding time dimension is 20 minutes. Next, determine the target curve segment from the current time point to the end time point on the power output curve. The target curve segment reflects the expected power output change. Then, divide this target curve segment into three power intervals: low, medium, and high. The division criteria can be based on the percentage of power values, such as defining 0-40% of the maximum power as a low power interval, 40-70% as a medium power interval, and 70-100% as a high power interval.

[0246] After that, calculate the power average of the low-power interval of the target curve segment and obtain the time period. First, collect all the power data points in the previously divided low-power interval. Assuming that the low-power interval contains 100 data points, and the power values ​​are 5kW, 7kW, 6kW, etc., then the average value is calculated by adding these power values ​​and dividing them by the number of data points. For example, if the sum of the power values ​​of these 100 points is 800kW, then the average power is 8kW. At the same time, record the duration of this low-power interval. If each data point represents 1 second, then the time period of this low-power interval is 100 seconds, that is, 1 minute and 40 seconds.

[0247] For the medium power interval, similar to S932, first determine all the data points in the medium power interval. Assume that the medium power interval contains 150 data points, and the power values ​​are 25kW, 28kW, 30kW, etc. When calculating the average, add these power values ​​and divide them by the number of data points. For example, if the sum of the power values ​​of 150 points is 4500kW, then the average power is 30kW. At the same time, record the duration of this medium power interval. If each data point represents 1 second, then the time period of the medium power interval is 150 seconds, that is, 2 minutes and 30 seconds.

[0248] For the high power interval, similar to S932 and S933, first determine all the data points in the high power interval. Assume that the high power interval contains 50 data points, and the power values ​​are 70kW, 75kW, 80kW, etc. When calculating the average value, add these power values ​​and divide them by the number of data points. For example, if the sum of the power values ​​of 50 points is 3750kW, then the average power is 75kW. At the same time, record the duration of this high power interval. If each data point represents 1 second, then the time period of the high power interval is 50 seconds. The high power interval usually corresponds to the acceleration or climbing state of the vehicle. Although these states are short in duration, they have high energy consumption and have a significant impact on the total energy consumption.

[0249] By querying the preset power output mode database, the energy conversion efficiency in different modes is obtained. The power output mode database stores the energy conversion efficiency data under different power output modes. Usually, electric vehicles have multiple driving modes, such as energy-saving mode, standard mode and power mode. In each mode, the energy conversion efficiency of the vehicle is different. For example, the energy-saving mode may have an energy conversion efficiency of 90%, the standard mode is 85%, and the power mode is 80%, reflecting the energy utilization efficiency of the vehicle under different operating conditions.

[0250] Multiply the average power value calculated previously by the corresponding time period to obtain the energy consumption of each power interval. Specifically, for the low power interval, if the average power is 8kW and the time period is 100 seconds (about 1.67 minutes), the energy consumption is calculated as: 8kW×(100 / 3600)h=0.22kWh. Similarly, for the medium power interval, if the average power is 30kW and the time period is 150 seconds (2.5 minutes), the energy consumption is calculated as: 30kW×(150 / 3600)h=1.25kWh. For the high power interval, if the average power is 75kW and the time period is 50 seconds (about 0.83 minutes), the energy consumption is calculated as: 75kW×(50 / 3600)h=1.04kWh.

[0251] Combine the previously calculated energy consumption of each power interval with the corresponding energy conversion efficiency to obtain a more accurate energy consumption value. Specifically, for the low power interval (corresponding to the energy-saving mode), if the energy consumption is 0.22kWh and the energy conversion efficiency is 90%, the actual energy consumption is calculated as: 0.22kWh / 0.9=0.244kWh. For the medium power interval (corresponding to the standard mode), if the energy consumption is 1.25kWh and the energy conversion efficiency is 85%, the actual energy consumption is calculated as: 1.25kWh / 0.85=1.47kWh. For the high power interval (corresponding to the power mode), if the energy consumption is 1.04kWh and the energy conversion efficiency is 80%, the actual energy consumption is calculated as: 1.04kWh / 0.8=1.3kWh. By considering the loss of energy conversion under different working conditions, the energy consumption estimate is closer to the actual situation.

[0252] Calculate and display the total energy consumption and energy savings of the electric public vehicle for the remaining mileage of the current route. First, add the actual energy consumption values ​​of the three power intervals calculated previously to obtain the total energy consumption. For example, if the energy consumption in the low power interval is 0.244kWh, the energy consumption in the medium power interval is 1.47kWh, and the energy consumption in the high power interval is 1.3kWh, then the total energy consumption is 0.244+1.47+1.3=3.014kWh. The total energy consumption represents the total amount of electricity that the vehicle is expected to consume in the remaining mileage. Next, compare the optimized energy consumption value with the energy consumption value previously predicted based on the energy consumption trend to calculate the energy savings. Assuming the energy consumption value predicted by the trend is 3.5kWh, the energy savings is 3.5kWh-3.014kWh=0.486kWh. Finally, display the energy savings.

[0253] This implementation achieves a high-precision estimation of the energy consumption of electric public vehicles by considering the energy conversion efficiency under different driving modes. Not only the energy consumption characteristics of the vehicle under different driving conditions are considered, but also the energy conversion efficiency is introduced, making the energy consumption prediction closer to the actual situation. By comparing the optimized energy consumption with the trend predicted energy consumption, the energy-saving effect is intuitively demonstrated, effectively improving the vehicle operation efficiency.

[0254] The present application also provides an electronic device, including:

[0255] a memory configured to store instructions; and

[0256] The processor is configured to call instructions from the memory and implement the above-mentioned public vehicle operation management method based on big data analysis when executing the instructions.

[0257] Figure 5 The structure diagram of a public vehicle operation management system based on big data analysis provided by an embodiment of the present application is shown as follows: Figure 5As shown, the embodiment of the present application also provides a public vehicle operation management system based on big data analysis, including:

[0258] Electronic device 10;

[0259] The vehicle terminal 20 of the electric bus is connected to the electronic device 10 .

[0260] An embodiment of the present application also provides a machine-readable storage medium, on which instructions are stored, and the instructions are used to enable a machine to execute the above-mentioned public vehicle operation management method based on big data analysis.

[0261] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0262] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0263] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0264] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0265] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0266] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0267] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0268] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0269] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. A public vehicle operation management method based on big data analysis, characterized in that: Applied to an electronic device, the electronic device is deployed with a big data model, and the electronic device is connected to a vehicle terminal of an electric public vehicle. The method includes: In response to receiving a battery alarm signal of the electric public vehicle, obtaining battery parameters and vehicle operation data of the electric public vehicle, where the vehicle operation data includes real-time road condition data, vehicle operation data, and driving behavior data of the electric public vehicle on the current driving route; and Obtaining current environmental data; Inputting the current environmental data and the real-time road condition data into the big data model to obtain the current vehicle operation scenario, where the big data model is constructed based on historical environmental data and historical road condition data; In a preset battery parameter database, querying to obtain a battery parameter range corresponding to the current vehicle operation scenario; When the battery parameters are within the battery parameter range, obtaining vehicle energy consumption data of the electric public vehicle within a preset kilometer range, and predicting an energy consumption trend of the electric public vehicle within a future preset time period according to the battery parameters, the vehicle energy consumption data, and the current vehicle operation scenario; Generating a power output mode of the electric public vehicle on the current driving route according to the energy consumption trend, the vehicle operation data, and the driving behavior data, where the power output mode is used to control the power distribution of the electric public vehicle; Generating a braking adjustment instruction according to the power output mode, where the braking adjustment instruction is used to instruct the electric public vehicle to execute a braking strategy; Sending the braking adjustment instruction to the vehicle terminal so that the vehicle terminal executes the braking strategy.

2. The method according to claim 1, characterized in that The construction steps of the big data model include: Using a preset feature extraction algorithm to extract features from the historical environmental data and the historical road condition data to obtain n first features, where n is an integer greater than 3; Using a first preset clustering algorithm to cluster the n first features to obtain m clustering clusters, where 0 < m < n and m is an integer; Using the feature extraction algorithm to extract features from the m clustering clusters to obtain p second features, where 0 < p < m and p is an integer; Using a second preset clustering algorithm to cluster the p second features to obtain q clustering results, where each clustering result includes a clustering center, and 0 < q < p and q is an integer; Using a preset distance algorithm to calculate the average distance between all the second features and each clustering center for each clustering center, and taking the clustering center corresponding to the smallest average distance as the target clustering center, and taking the clustering result corresponding to the target clustering center as the target clustering result; Mapping the target clustering result to the corresponding operation scenario according to a preset mapping function; Encapsulating the feature extraction algorithm, the first preset clustering algorithm, the second preset clustering algorithm, the distance algorithm, and the mapping function to obtain the big data model.

3. The method according to claim 1, characterized in that The method further includes: When the battery parameter is not within the battery parameter range, obtaining the current power of the electric public vehicle through the vehicle terminal, and calculating the maximum driving distance of the electric public vehicle according to the current power; Obtaining the current location of the electric public vehicle; Determining the remaining driving distance of the electric public vehicle on the current driving route according to the current positioning; When the remaining driving distance is greater than the maximum driving distance, calculating the straight-line distance between the current location and the end point of the current driving route, and searching for an electric public vehicle charging station in an area with the current location as the center and the straight-line distance as the radius; In the case where there are multiple electric public vehicle charging stations, taking the electric public vehicle charging station that is shortest from the current location as the target charging station; A new driving route is generated according to the current location, the target charging station and the end point of the current driving route, and the new driving route is sent to the vehicle terminal.

4. The method according to claim 1, characterized in that: The predicting of the energy consumption trend of the electric public vehicle within a preset time period in the future according to the battery parameters, the vehicle energy consumption data and the current operation scenario of the vehicle includes: The battery parameters, the vehicle energy consumption data and the current operation scenario of the vehicle are input into a pre-trained prediction model to obtain the energy consumption trend of the electric public vehicle within a future preset time period.

5. The method according to claim 1, characterized in that The step of generating a power output mode of the electric public vehicle on the current driving route according to the energy consumption trend, the vehicle operation data and the driving behavior data comprises: According to the energy consumption trend, the power supply corresponding to each time step is calculated, and the power supply corresponding to each time step is fitted to generate a power supply curve of the electric public vehicle; Determining the driving state of the electric public vehicle according to the vehicle operation data, wherein the driving state includes an idle state, an acceleration state, a constant speed state, and a deceleration state; Acquiring an energy conversion efficiency curve of the electric public vehicle, and determining an optimal output power of the electric public vehicle in the driving state according to the power supply power curve and the energy conversion efficiency curve; Calculating the curvature of the power supply curve in real time; When the curvature is greater than a preset threshold, a power output mode is determined according to the driving behavior data and the optimal output power of the electric public vehicle in the driving state.

6. The method according to claim 5, characterized in that The step of determining the optimal output power of the electric public vehicle in the driving state according to the power supply power curve and the energy conversion efficiency curve comprises: According to the power supply power curve and the energy conversion efficiency curve, obtaining a power supply power data segment and a corresponding energy conversion efficiency data segment in the driving state; For each time point of the power supply data segment and the corresponding energy conversion efficiency data segment, calculating the actual output power at each time point in the driving state, and generating an output power curve according to the actual output power at each time point; According to the output power curve and the energy conversion efficiency curve, an optimal efficiency range is determined, and an average value of the output power corresponding to the optimal efficiency range is calculated, and the average value is used as the optimal output power of the electric public vehicle in the driving state.

7. The method according to claim 6, characterized in that The determining of the power output mode according to the driving behavior data and the optimal output power of the electric public vehicle in the driving state includes: Determining a driving behavior type according to the driving behavior data, wherein the driving behavior type includes smooth, medium and aggressive; Acquire a weight coefficient corresponding to the driving behavior type in a preset database; Calculating a power adjustment coefficient of the actual output power corresponding to each time point according to the weight coefficient, wherein the product of the optimal output power and the power adjustment coefficient is less than the maximum output power of the electric public vehicle; Calculating the product of the actual output power at each time point and the corresponding power adjustment coefficient to obtain a plurality of adjusted output power values, and generating an optimized power output curve according to all processed output power values; Determining a power output mode according to the optimal output power and the optimized power output curve; Among them, the power output curve includes a low power interval, a medium power interval and a high power interval, and the power output mode includes an energy-saving mode, a standard mode and a power mode. When the optimal output power is located in the low power interval of the power output curve, the power output mode is the energy-saving mode, when the optimal output power is located in the medium power interval of the power output curve, the power output mode is the standard mode, and when the optimal output power is located in the high power interval of the power output curve, the power output mode is the power mode.

8. The method according to claim 7, characterized in that The driving behavior data includes the frequency of sudden acceleration and sudden braking, the vehicle turning angle, the vehicle turning speed, and the vehicle idling time. Determining the driving behavior type based on the driving behavior data includes: Extracting features of the frequency of sudden acceleration and sudden braking, the vehicle turning angle, the vehicle turning speed, and the vehicle idling time to obtain a plurality of driving behavior features; The driving behavior characteristics are input into a pre-trained classification model to obtain the driving behavior type.

9. An electronic device, characterized in that: include: a memory configured to store instructions; as well as A processor is configured to call the instructions from the memory and to implement the public vehicle operation management method based on big data analysis according to any one of claims 1 to 8 when executing the instructions.

10. A public vehicle operation management system based on big data analysis, characterized in that: include: An electronic device according to claim 9; The vehicle terminal of the electric public vehicle is connected to the electronic device.

Images