Unmanned intelligent analysis method and system

Abstract

The invention relates to the technical field of driving analysis, in particular to an unmanned intelligent analysis method and system, and the method comprises the following steps: S1, collecting multi-dimensional data in real time through a vehicle-mounted sensor, a traffic information network and meteorological data, building a short-term energy consumption prediction model based on the multi-dimensional data, the short-term energy consumption prediction model predicts energy consumption fluctuation in a future driving route; s2, dynamically adjusting a driving mode of the unmanned vehicle according to an energy consumption fluctuation prediction result of the short-term energy consumption prediction model and based on a multi-mode energy efficiency optimization mechanism; and S3, through a self-adaptive friction coefficient feedback mechanism, road surface friction coefficient changes of the tires and the ground under different road conditions are dynamically evaluated, and power output and braking response of the vehicle are adjusted. Acceleration, deceleration, sliding and regenerative braking modes of the vehicle can be adjusted in real time, and it is ensured that the vehicle is in the optimal energy efficiency state all the time.

Description

Technical Field

[0001] The present invention relates to the technical field of driving analysis, and in particular to an unmanned driving intelligent analysis method and system. Background Art

[0002] With the rapid development of driverless technology, autonomous vehicles have gradually become a core component of intelligent transportation systems. Existing driverless systems usually rely on on-board sensors (such as lidar, cameras, millimeter-wave radar, etc.) to collect data on the surrounding environment, and combine deep learning algorithms for environmental perception and path planning. However, the existing technology still has many shortcomings in complex dynamic environments, especially when dealing with low-friction roads such as wet, icy, and gravel. The vehicle's power output and braking response are often not accurate enough, which can easily cause the vehicle to slip or lose control. In addition, how to optimize energy consumption management and improve endurance under different road conditions and changing climate conditions is also a technical problem that needs to be urgently solved in the field of driverless driving.

[0003] In existing unmanned driving systems, energy consumption management and driving mode adjustments are mostly based on preset rules and lack the ability to intelligently respond to real-time data changes. For example, when the road friction coefficient changes, it is impossible to adaptively adjust the power output and braking response according to different road conditions, resulting in reduced driving safety, especially in severe weather conditions. In addition, traditional energy efficiency management methods cannot effectively handle the impact of multi-dimensional data (such as vehicle speed, slope, friction coefficient, climatic conditions, etc.) on energy consumption, and lack accurate energy consumption prediction and real-time adjustment mechanisms, which further limits the overall energy efficiency optimization of unmanned vehicles. Summary of the invention

[0004] The present invention provides an unmanned driving intelligent analysis method and system.

[0005] An unmanned driving intelligent analysis method comprises the following steps: S1, multi-dimensional data collection and prediction: Through on-board sensors, traffic information networks and meteorological data, multi-dimensional data is collected in real time, including road conditions, slope, road friction coefficient, weather conditions (temperature, wind speed, humidity), and a short-term energy consumption prediction model is built based on the multi-dimensional data. The short-term energy consumption prediction model predicts energy consumption fluctuations in future driving routes; S2, real-time driving mode dynamic adjustment: According to the energy consumption fluctuation prediction results of the S1 medium- and short-term energy consumption prediction model, and based on the multi-modal energy efficiency optimization mechanism, the driving mode of the unmanned vehicle is dynamically adjusted, including acceleration mode, deceleration mode, gliding mode, and regenerative braking mode; the dynamic adjustment not only considers the current energy consumption, but also combines the vehicle battery status, current speed and future road conditions (slope conditions) to optimize the vehicle's driving status in real time; S3, adaptive road friction feedback and driving strategy optimization: Based on the real-time road friction coefficient collected in S1, the adaptive friction coefficient feedback mechanism dynamically evaluates the changes in the road friction coefficient between the tire and the ground under different road conditions, and adjusts the vehicle's power output and braking response to ensure optimal energy efficiency and driving safety in wet, icy or loose road conditions.

[0006] Optionally, the multi-dimensional data collection of S1 includes: On-board sensor data collection: On-board sensors, including lidar, camera, millimeter-wave radar, and inertial measurement unit, collect road condition information and slope data of the vehicle's surrounding environment in real time; lidar is used to measure road obstacles and road morphology, and the inertial measurement unit captures the vehicle's tilt angle and acceleration changes to detect slope changes; Traffic information network data acquisition: through the connection with the external traffic information network, the current road traffic flow, congestion, accident information and road construction conditions can be obtained in real time; Weather condition collection and analysis: Through data connection with the meteorological network, the current weather conditions of the environment are obtained in real time, including temperature, wind speed and humidity parameters. The weather conditions are used to analyze the weather conditions that affect vehicle driving; Real-time estimation of road friction coefficient: Through the data analysis of the on-board wheel speed sensor and the braking system, combined with real-time road feedback (such as the activation frequency of the ABS anti-lock braking system), the current road friction coefficient is inferred and cross-validated with the historical data of the traffic network to ensure the accuracy of the friction coefficient estimation. The real-time estimation of the road friction coefficient is expressed as: ,in, is the road friction coefficient estimated in real time. is the traction between the vehicle and the ground (or the friction between the tires and the road), is the vehicle mass, is the acceleration due to gravity.

[0007] Optionally, the construction of the short-term energy consumption prediction model in S1 specifically includes: The collected multi-dimensional data is used as characteristic data related to energy consumption, including road conditions, slope, road friction coefficient and weather conditions; A short-term energy consumption prediction model is constructed using a multi-layer feedforward neural network. By inputting feature data, the model is trained to predict energy consumption fluctuations in future driving routes. The short-term energy consumption prediction model is continuously iterated and updated based on historical data and real-time collected data to improve prediction accuracy. Based on the trained short-term energy consumption prediction model, real-time data of the current road section is input to predict energy consumption fluctuations in future driving routes, and energy consumption prediction results for each road section are output.

[0008] Optionally, the input characteristic data of the short-term energy consumption prediction model is expressed as , including the following parameters: : The vehicle at time speed at 1000 s; : The vehicle at time The acceleration at : The vehicle at time The slope angle of the road; : Road friction coefficient, at time Real-time estimation of :time The ambient temperature at :time Ambient humidity at the time; :time Wind speed at 1000 km / h; After preprocessing, the feature data is used as the input feature of the multi-layer feedforward neural network: ; The model structure of the short-term energy consumption prediction model includes: Input layer: The input layer receives Vector with dimension , that is, the number of input features, the input dimension ; Hidden layer: It is the key part of the model to handle complex nonlinear relationships. Each hidden layer consists of multiple neurons. hidden layers, and the number of neurons in each layer is ,in , the calculation formula for each layer is: ,in, For the The output of the hidden layer, For the The weight matrix of the layer has the dimension , represents the weight between the current layer and the previous layer of neurons, For the The bias term of the layer has the dimension , ReLU is the activation function, defined as: ; The number of neurons in each layer Set according to task complexity; Output layer: The output layer predicts the energy consumption value of the vehicle at the current moment , the output of this layer is calculated by a linear activation function: ,in, is the weight matrix of the output layer, with dimension , maps the output of the hidden layer to a single energy consumption value, is the bias term of the output layer, with a dimension of 1×1, For time The predicted value of energy consumption at When training the model, minimize the energy consumption prediction error and use the mean square error as the loss function; After the model training is completed, enter the current time The eigenvector of , calculate the energy consumption value In order to predict the energy consumption fluctuations in the future driving route, a sliding window is used to predict the future Energy consumption at each moment: ; in, It is the input feature of the future road section. By inputting the multi-dimensional data of the future, the energy consumption value at different time points is predicted. It is a multi-layer feedforward neural network. Especially under conditions such as uphill, downhill, slippery roads or strong winds, the fluctuation of energy consumption is more significant.

[0009] Optionally, the real-time dynamic adjustment of the driving mode in S2 specifically includes: S21, based on the short-term energy consumption prediction model, real-time analysis of energy consumption fluctuations , select the appropriate driving mode, and calculate the energy consumption fluctuation based on the predicted value of energy consumption at adjacent moments , driving mode selection is based on energy consumption fluctuations ,when When the energy consumption increases, consider the deceleration mode or the gliding mode. When it is on, it means that the energy consumption is reduced, the current driving mode is maintained, or it is adjusted according to the battery status; S22, battery status perception and optimization adjustment: Battery status perception is achieved by combining the battery power status Analyze the relationship between battery power status and energy consumption to maintain energy efficiency balance by optimizing driving mode in real time. When the vehicle is in a state of slowing down, the glide mode or regenerative braking mode is preferred to extend the driving range. When , the acceleration mode is allowed; S23, optimization and adjustment of future road conditions and driving status: In future road conditions, including the impact of slope changes on vehicle energy consumption, the calculation formula is as follows: ,in, The vehicle is at a slope angle The power consumption under is the vehicle mass, is the acceleration due to gravity, For time The speed of the vehicle at For time The slope angle of the road; when the vehicle encounters an uphill slope, the slope Increase, power consumption The vehicle speed also increases accordingly, and the deceleration mode is selected first; when the vehicle encounters a downhill slope, the coasting mode is selected to reduce energy consumption; S24, energy efficiency optimization feedback adjustment mechanism: In the energy efficiency optimization feedback mechanism, dynamic adjustments are made according to the real-time state and energy consumption of the vehicle. The energy efficiency is calculated as follows: ,in, For time The energy efficiency is the speed per unit of energy consumption. is the instantaneous speed of the vehicle, For time By monitoring energy efficiency , real-time switching between driving modes, when When the vehicle speed is lower than the preset threshold, it indicates low energy efficiency and the vehicle will be decelerated or coasted to improve energy efficiency. When the preset threshold is exceeded, it indicates that the vehicle is energy-efficient in the current mode and maintains the current driving state.

[0010] Optionally, the real-time driving mode dynamic adjustment further includes preset adjustment priorities, and the adjustment priorities are arranged as follows: Battery status perception and optimization adjustment > Future road conditions and driving status optimization adjustment > Energy consumption fluctuation adjustment > Energy efficiency optimization feedback adjustment mechanism.

[0011] Optionally, the S3 specifically includes: S31, friction coefficient based on real-time acquisition The adaptive feedback mechanism dynamically evaluates the friction changes under different road conditions and calculates the rate of change of the friction coefficient. To identify rapid fluctuations in friction, it is calculated as: ,according to The size and trend of the vehicle's braking force can be used to determine changes in current road conditions (such as a slippery road suddenly becoming more slippery, or an icy road returning to normal conditions), providing feedback information for the power output and braking systems; S32, based on friction coefficient and its rate of change , adjust vehicle power output in real time When the friction coefficient is reduced (such as on wet or icy roads), reduce the vehicle's power output to reduce the risk of wheel slippage; when the friction is restored, gradually increase the power output to restore normal driving conditions. S33, when the friction coefficient is low or the road conditions are complex, the Dynamically adjust the vehicle's braking response When the friction force decreases, the feedback response of the braking system is increased to avoid excessive braking force causing wheel lock or loss of control; when the friction coefficient increases, the braking feedback is reduced to ensure the smoothness of the braking operation. The formula for adjusting the braking response is: , ensuring that the braking response is more sensitive when the friction coefficient is low, preventing excessive braking force; when the friction coefficient is high, the system allows greater braking force.

[0012] Optionally, the vehicle power output adjustment in S32 is expressed as: ,in, is the adjustment factor, is the current road friction coefficient. As the friction coefficient decreases, the power output decreases proportionally, and vice versa, it gradually returns to normal.

[0013] Optionally, the adjustment factor According to the friction coefficient of the road The change range and current speed of the vehicle , and wheel slip to dynamically adjust the adjustment coefficient Calculated as: ,in, is the rate of change of the friction coefficient, which reflects the degree of mutation of the road condition. is the current speed of the vehicle, It is the wheel slip detection parameter, which reflects the wheel slip situation through the slip degree of the wheel contacting with the ground (through the wheel slip rate or ABS system feedback). When the slip increases (i.e., the slip is severe), the adjustment coefficient should be increased to reduce the power output or adjust the braking response more quickly. , , It is an empirical weight coefficient determined through experimental data to balance the effects of changes in friction coefficient, speed, and slippage on the adjustment range.

[0014] An unmanned driving intelligent analysis system, used to implement the above-mentioned unmanned driving intelligent analysis method, includes the following modules: Multi-dimensional data acquisition module, used to collect multi-dimensional data in real time through vehicle sensors, traffic information network and meteorological data, including vehicle speed, acceleration, slope, road friction coefficient, weather conditions; A short-term energy consumption prediction module, which constructs a short-term energy consumption prediction model based on the multi-dimensional data to predict energy consumption fluctuations in future driving routes; A driving mode dynamic adjustment module dynamically adjusts the driving mode, including acceleration mode, deceleration mode, gliding mode and regenerative braking mode, according to the energy consumption prediction result, the battery status of the vehicle and the future road conditions, so as to optimize the energy efficiency and driving safety of the vehicle; Adaptive friction coefficient feedback module, which evaluates the change of road friction coefficient in real time during vehicle driving and dynamically adjusts the vehicle's power output and braking response according to the change of friction coefficient; The priority management module is used to set priorities among energy consumption fluctuations, road condition changes, energy efficiency optimization feedback adjustment mechanism and battery status perception to avoid driving mode adjustment conflicts and ensure the best performance and stability of the system in complex environments.

[0015] Beneficial effects of the present invention: The present invention introduces a multi-source data collection and real-time prediction mechanism, combines on-board sensors, traffic information networks and meteorological data to build a short-term energy consumption prediction model, and accurately predicts energy consumption fluctuations in future driving routes by dynamically analyzing road conditions, slopes and climate conditions. It also dynamically adjusts the driving mode based on a multimodal energy efficiency optimization mechanism. This mechanism improves the energy efficiency management of vehicles in complex environments, minimizes unnecessary energy consumption, and achieves efficient energy utilization during driving. Compared with the prior art, the present invention can adjust the acceleration, deceleration, gliding and regenerative braking modes of the vehicle in real time to ensure that the vehicle is always in the optimal energy efficiency state.

[0016] The present invention can dynamically evaluate the change of the friction coefficient between the tire and the ground under different road conditions by collecting the road friction coefficient data in real time and combining it with an adaptive friction coefficient feedback mechanism. According to the real-time change rate of the friction coefficient, the current speed of the vehicle and the slipping situation, the system adaptively adjusts the power output and braking response to ensure the driving safety of the vehicle under complex road conditions such as wet, icy or loose roads. Compared with the fixed response mechanism of the traditional system, the present invention realizes intelligent regulation of power and braking by adjusting the coefficient α, especially when slipping or the friction coefficient suddenly changes, the power output and braking force can be quickly adjusted to avoid vehicle loss of control and improve the overall driving stability.

[0017] The present invention solves the potential conflicts among energy consumption fluctuation analysis, future road condition prediction, battery status perception and energy efficiency feedback mechanism by introducing a priority adjustment mechanism, gives priority to processing battery status and future road condition changes, ensures that the vehicle prioritizes safe and efficient driving modes when the battery is low or the road conditions with large slope changes, avoids the conflict problem when adjusting the driving mode due to multiple factors at the same time, and through dynamic priority adjustment, can comprehensively optimize the driving mode according to actual driving conditions to ensure a balance among endurance, driving safety and energy efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0019] Figure 1 A schematic diagram of a method flow chart of an embodiment of the present invention; Figure 2 Schematic diagram of system function modules according to an embodiment of the present invention. DETAILED DESCRIPTION

[0020] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. At the same time, it is explained here that in order to make the embodiments more detailed, the following embodiments are the best and preferred embodiments, and those skilled in the art may also adopt other alternatives to implement some known technologies; and the accompanying drawings are only for more specific description of the embodiments, and are not intended to specifically limit the present invention.

[0021] It should be noted that the references to "one embodiment", "embodiment", "exemplary embodiments", "some embodiments" and the like in the specification indicate that the embodiments described may include specific features, structures or characteristics, but not every embodiment may include the specific features, structures or characteristics. In addition, when a specific feature, structure or characteristic is described in conjunction with an embodiment, it should be within the knowledge of a person skilled in the art to implement such feature, structure or characteristic in conjunction with other embodiments (whether or not explicitly described).

[0022] In general, a term can be understood, at least in part, from its use in context. For example, depending, at least in part, on the context, the term "one or more" as used herein can be used to describe any feature, structure, or characteristic in the singular sense, or can be used to describe a combination of features, structures, or characteristics in the plural sense. Additionally, the term "based on" can be understood as not necessarily intended to convey an exclusive set of factors, but can instead, depending, at least in part, on the context, allow for the presence of other factors that are not necessarily explicitly described.

[0023] like Figure 1 As shown, an unmanned driving intelligent analysis method realizes the endurance optimization and energy consumption management of unmanned vehicles in a changing environment by introducing road condition perception, real-time climate analysis and multi-modal energy efficiency optimization mechanism, including the following steps: S1, multi-dimensional data collection and prediction: Through on-board sensors, traffic information networks and meteorological data, multi-dimensional data is collected in real time, including road conditions, slope, road friction coefficient, weather conditions (temperature, wind speed, humidity), and a short-term energy consumption prediction model is built based on the multi-dimensional data. The short-term energy consumption prediction model predicts energy consumption fluctuations in future driving routes; S2, real-time driving mode dynamic adjustment: According to the energy consumption fluctuation prediction results of the S1 medium- and short-term energy consumption prediction model, and based on the multi-modal energy efficiency optimization mechanism, the driving mode of the unmanned vehicle is dynamically adjusted, including acceleration mode, deceleration mode, gliding mode, and regenerative braking mode; the dynamic adjustment not only considers the current energy consumption, but also combines the vehicle battery status, current speed and future road conditions (slope conditions) to optimize the vehicle's driving status in real time; S3, adaptive road friction feedback and driving strategy optimization: Based on the real-time road friction coefficient collected in S1, the adaptive friction coefficient feedback mechanism dynamically evaluates the changes in the road friction coefficient between the tire and the ground under different road conditions, adjusts the vehicle's power output and braking response to ensure optimal energy efficiency and driving safety in slippery, icy or loose road conditions, and identifies possible low-friction sections in the future in advance and adjusts driving strategies to avoid increased energy consumption and driving safety risks caused by changes in friction.

[0024] The goal of S2 is to adjust the driving mode according to future energy consumption predictions through a multimodal energy efficiency optimization mechanism, while the goal of S3 is to optimize energy efficiency and safety by dynamically adjusting power output and braking under different friction conditions. S3 focuses more on detailed adjustments under specific conditions (wet or icy roads), which is a further optimization based on the S2 driving mode adjustment.

[0025] The driving mode adjustment of S2 mainly focuses on the dynamic changes of the vehicle's overall driving state, while S3 provides an additional feedback mechanism when the specific friction changes. If the vehicle enters a low-friction section after S2 adjusts the driving mode, S3 will further adjust the power output and braking to ensure energy efficiency and safety in this specific environment. This shows that S3 is a supplement to S2, achieving more refined optimization based on the driving mode adjustment.

[0026] S1's multi-dimensional data collection includes: On-board sensor data collection: On-board sensors, including lidar, camera, millimeter-wave radar, and inertial measurement unit, collect road condition information and slope data of the vehicle's surrounding environment in real time; lidar is used to measure road obstacles and road morphology, and the inertial measurement unit captures the vehicle's tilt angle and acceleration changes to detect slope changes; Traffic information network data acquisition: Through the connection with the external traffic information network, the current road traffic flow, congestion, accident information and road construction conditions can be obtained in real time. The traffic information network data can be dynamically updated to ensure that the vehicle can optimize the driving strategy according to the latest road condition information; Weather condition collection and analysis: Through data connection with the meteorological network, the current weather conditions of the environment are obtained in real time, including temperature, wind speed and humidity parameters. The weather conditions are used to analyze the weather conditions that affect vehicle driving. At the same time, the vehicle's built-in humidity sensor and temperature sensor will also detect the microclimate environment around the vehicle in real time to ensure accurate judgment of the local climate; Real-time estimation of road friction coefficient: Through the data analysis of the on-board wheel speed sensor and the braking system, combined with real-time road feedback (such as the activation frequency of the ABS anti-lock braking system), the current road friction coefficient is inferred and cross-validated with the historical data of the traffic network to ensure the accuracy of the friction coefficient estimation. The real-time estimation of the road friction coefficient is expressed as: ,in, is the road friction coefficient estimated in real time. is the traction between the vehicle and the ground (or the friction between the tires and the road), is the vehicle mass, is the acceleration due to gravity; Traction Calculation of traction: The actual friction is estimated by the vehicle's power output and speed changes, especially when the vehicle is accelerating, decelerating or braking. The sensor can monitor the wheel slippage and wheel speed differences to estimate the actual friction: ,in, The acceleration or deceleration of the vehicle can be measured in real time by the on-board inertial measurement unit (IMU).

[0027] The construction of S1 short- and medium-term energy consumption prediction model specifically includes: The collected multi-dimensional data is used as characteristic data related to energy consumption, including road conditions, slope, road friction coefficient and weather conditions; A short-term energy consumption prediction model is constructed using a multi-layer feedforward neural network. By inputting feature data, the model is trained to predict energy consumption fluctuations in future driving routes. The short-term energy consumption prediction model is continuously iterated and updated based on historical data and real-time collected data to improve prediction accuracy. Based on the trained short-term energy consumption prediction model, real-time data of the current road section is input to predict energy consumption fluctuations in future driving routes, and energy consumption prediction results for each road section are output, especially for significant fluctuations in energy consumption under uphill, downhill, slippery or severe weather conditions, to generate a future energy consumption fluctuation map for the vehicle to dynamically adjust its driving strategy.

[0028] The input characteristic data of the short-term energy consumption prediction model is expressed as , including the following parameters: : The vehicle at time speed at 1000 s; : The vehicle at time The acceleration at : The vehicle at time The slope angle of the road; : Road friction coefficient, at time Real-time estimation of :time The ambient temperature at :time Ambient humidity at the time; :time Wind speed at 1000 km / h; After preprocessing, the feature data is used as the input feature of the multi-layer feedforward neural network: ; The model structure of the short-term energy consumption prediction model includes: Input layer: The input layer receives Vector with dimension , that is, the number of input features, the input dimension , representing 7 input features: speed, acceleration, slope, friction coefficient, temperature, humidity and wind speed, ; Hidden layer: It is the key part of the model to handle complex nonlinear relationships. Each hidden layer consists of multiple neurons. hidden layers, and the number of neurons in each layer is ,in , the calculation formula for each layer is: ,in, For the The output of the hidden layer, For the The weight matrix of the layer has the dimension , represents the weight between the current layer and the previous layer of neurons, For the The bias term of the layer has the dimension , ReLU is the activation function, defined as: ; The number of neurons in each layer Set according to task complexity, for example: First hidden layer , the second hidden layer , the third hidden layer ; Output layer: The output layer predicts the energy consumption value of the vehicle at the current moment , the output of this layer is calculated by a linear activation function: ,in, is the weight matrix of the output layer, with dimension , maps the output of the hidden layer to a single energy consumption value, is the bias term of the output layer, with a dimension of 1×1, For time The predicted value of energy consumption at When training the model, the energy consumption prediction error is minimized and the mean square error is used as the loss function, which is defined as: ,in, For the model at time The predicted energy consumption value, For time The actual energy consumption value, is the number of training samples.

[0029] After the model training is completed, enter the current time The eigenvector of , calculate the energy consumption value In order to predict the energy consumption fluctuations in the future driving route, a sliding window is used to predict the future Energy consumption at each moment: ; in, It is the input feature of the future road section. By inputting the multi-dimensional data of the future, the energy consumption value at different time points is predicted. It is a multi-layer feedforward neural network. Especially under conditions such as uphill, downhill, slippery roads or strong winds, the fluctuation of energy consumption is more significant.

[0030] The real-time driving mode dynamic adjustment in S2 includes: S21, based on the short-term energy consumption prediction model, real-time analysis of energy consumption fluctuations , select the appropriate driving mode, and calculate the energy consumption fluctuation based on the predicted value of energy consumption at adjacent moments , expressed as: ,in, For in time The predicted energy consumption value, is the energy consumption value at the previous moment, Energy consumption fluctuation, indicating the difference in energy consumption between two moments. The driving mode is selected based on energy consumption fluctuation. ,when When the energy consumption increases, consider the deceleration mode or the gliding mode. When it is on, it means that the energy consumption is reduced, the current driving mode is maintained, or it is adjusted according to the battery status; S22, battery status perception and optimization adjustment: Battery status perception is achieved by combining the battery power status Analysis, the calculation formula of battery status: ,in, For time The remaining battery power, For time The remaining battery power The total capacity of the battery, the relationship between the battery charge state and energy consumption is maintained by optimizing the driving mode in real time. When the vehicle is in a state of slowing down, the glide mode or regenerative braking mode is preferred to extend the driving range. When , the acceleration mode is allowed; S23, optimization and adjustment of future road conditions and driving status: In future road conditions, including the impact of slope changes on vehicle energy consumption, the calculation formula is as follows: ,in, The vehicle is at a slope angle The power consumption under is the vehicle mass, is the acceleration due to gravity, For time The speed of the vehicle at For time The slope angle of the road; when the vehicle encounters an uphill slope, the slope Increase, power consumption The vehicle speed also increases accordingly, and the deceleration mode is selected first; when the vehicle encounters a downhill slope, the coasting mode is selected to reduce energy consumption; S24, energy efficiency optimization feedback adjustment mechanism: In the energy efficiency optimization feedback mechanism, dynamic adjustments are made according to the real-time state and energy consumption of the vehicle. The energy efficiency is calculated as follows: ,in, For time The energy efficiency is the speed per unit of energy consumption. is the instantaneous speed of the vehicle, For time By monitoring energy efficiency , real-time switching between driving modes, when When the vehicle speed is lower than the preset threshold, it indicates low energy efficiency and the vehicle will be decelerated or coasted to improve energy efficiency. When the preset threshold is exceeded, it indicates that the vehicle is energy efficient in the current mode and maintains the current driving state; 1. Energy consumption fluctuation formula Used to analyze the changing trend of energy consumption and select the appropriate driving mode.

[0031] 2. Battery status By sensing the remaining battery power in real time, the driving mode is dynamically adjusted to maintain optimal energy efficiency.

[0032] 3. Slope power consumption formula Used to analyze the impact of slope on energy consumption in future road conditions and select the appropriate driving mode.

[0033] 4. Energy efficiency optimization formula By monitoring the vehicle's energy efficiency in real time and dynamically switching between different driving modes, energy efficiency is maximized.

[0034] Set priorities for different adjustment mechanisms, giving priority to the factors that have the greatest and most urgent impact on driving mode. The following is the priority scheme: Battery status awareness adjusts driving mode (highest priority): When the remaining battery power is below a certain threshold, endurance becomes the system's primary concern, so the system prioritizes energy-saving modes (such as coasting mode or regenerative braking mode) to extend endurance.

[0035] Priority processing: When the battery power is low (such as SOC(t) < 20%), the system will force the selection of coasting or regenerative braking mode to prevent the vehicle from being unable to continue driving.

[0036] Future road condition analysis to adjust driving mode (second priority): When encountering steep slopes or downhills, road conditions have a significant impact on vehicle energy consumption. Even if energy efficiency or energy consumption fluctuations are good, if future road conditions require adjustment (such as large changes in slope), the driving mode should be adjusted first to ensure a balance between driving safety and energy efficiency.

[0037] Priority processing: If the system detects that it is about to enter an uphill section, although the current energy efficiency is better, it is still necessary to select the deceleration mode to avoid excessive energy consumption; when going downhill, the coasting mode is selected to recover energy.

[0038] Energy consumption fluctuation analysis to adjust driving mode (medium priority): Analysis of energy consumption fluctuations can provide real-time optimization basis for adjusting driving mode, but its priority may be slightly lower than battery status and future road conditions, because these factors directly affect battery life and driving safety.

[0039] Priority processing: When the battery status and road conditions are normal, the system selects the appropriate mode, such as acceleration, deceleration or coasting, based on the energy consumption fluctuation range.

[0040] Energy efficiency optimization feedback mechanism (lowest priority): This mechanism selects the driving mode by real-time monitoring of energy efficiency and is a feedback-based adjustment. Since its main function is to fine-tune energy efficiency, it is usually given a lower priority, especially when there are special requirements for battery status or road conditions.

[0041] Priority processing: When the battery status, road conditions and energy consumption fluctuations are all within the normal range, the system dynamically switches driving modes based on the energy efficiency feedback mechanism.

[0042] In order to ensure that the system can respond flexibly in a variety of situations, a dynamic priority adjustment mechanism can be implemented: When the battery status is seriously below the safety threshold (such as SOC(t) < 10%), the system can temporarily ignore other optimization factors and give priority to ensuring battery life.

[0043] Under relatively flat road conditions, the system gives priority to energy consumption fluctuations and energy efficiency feedback to maximize energy utilization efficiency.

[0044] Real-time driving mode dynamic adjustment also includes preset adjustment priorities, which are ranked as follows: Battery status perception and optimization adjustment > Future road conditions and driving status optimization adjustment > Energy consumption fluctuation adjustment > Energy efficiency optimization feedback adjustment mechanism.

[0045] S3 specifically includes: S31, friction coefficient based on real-time acquisition The adaptive feedback mechanism dynamically evaluates the friction changes under different road conditions and calculates the rate of change of the friction coefficient. To identify rapid fluctuations in friction, it is calculated as: ,according to The size and trend of the vehicle's braking force can be used to determine changes in current road conditions (such as a slippery road suddenly becoming more slippery, or an icy road returning to normal conditions), providing feedback information for the power output and braking systems; S32, based on friction coefficient and its rate of change , adjust vehicle power output in real time When the friction coefficient is reduced (such as on wet or icy roads), reduce the vehicle's power output to reduce the risk of wheel slippage; when the friction is restored, gradually increase the power output to restore normal driving conditions. S33, when the friction coefficient is low or the road conditions are complex, the Dynamically adjust the vehicle's braking response When the friction force decreases, the feedback response of the braking system is increased to avoid excessive braking force causing wheel lock or loss of control; when the friction coefficient increases, the braking feedback is reduced to ensure the smoothness of the braking operation. The formula for adjusting the braking response is: , ensuring that the braking response is more sensitive when the friction coefficient is low, preventing excessive braking force; when the friction coefficient is high, the system allows greater braking force.

[0046] The vehicle power output adjustment in S32 is expressed as: ,in, is the adjustment factor, is the current road friction coefficient. As the friction coefficient decreases, the power output decreases proportionally, and vice versa, it gradually returns to normal.

[0047] Adjustment factor According to the friction coefficient of the road The change range and current speed of the vehicle , and wheel slip to dynamically adjust the adjustment coefficient Calculated as: ,in, is the rate of change of the friction coefficient, which reflects the degree of mutation of the road condition. The current speed of the vehicle. The higher the speed, the smaller the adjustment coefficient, in order to avoid instability caused by drastic adjustments at high speeds. It is the wheel slip detection parameter, which reflects the wheel slip situation through the slip degree of the wheel contacting with the ground (through the wheel slip rate or ABS system feedback). When the slip increases (i.e., the slip is severe), the adjustment coefficient should be increased to reduce the power output or adjust the braking response more quickly. , , It is an empirical weight coefficient determined by experimental data to balance the influence of friction coefficient change, speed, and slippage on the adjustment range; :Control the change of friction coefficient The more drastic the change in friction coefficient, the more the adjustment coefficient The larger the value, the faster the response time will be when the road condition changes suddenly (such as sudden slippery conditions). :Control speed The higher the speed, the smaller the adjustment coefficient should be to avoid instability caused by drastic power adjustment. On the contrary, at low speed, a larger adjustment coefficient can be allowed. : Reflects the impact of skidding. If the wheel skids severely, the system needs to quickly reduce power output or increase brake feedback. Should be enlarged.

[0048] Determine the weight coefficient through experimental data , , .

[0049] Experimental process: Experiment 1: Changes in friction coefficient under different road conditions: The test vehicle is driven on different road surfaces, such as wet, icy, and gravel roads, and different friction coefficients are recorded. The rate of change , and analyze the stability of the vehicle's acceleration, power output and braking response under different friction coefficients, and adjust the The value of ensures that the system can respond quickly when the road conditions change drastically.

[0050] Experiment 2: Stability test at different speeds The vehicle was tested at different speeds (20km / h, 50km / h, 80km / h) to adjust the power output and braking response, and observe the stability and driving comfort of the vehicle at different speeds. , ensuring that excessive power adjustments are avoided when driving at high speeds and reducing vehicle instability.

[0051] Experiment 3: Adjustment in case of wheel slip Simulate slip and record the wheel slip rate or the slip degree reported by the ABS system , and analyze the power output and brake feedback adjustment range under different slip conditions, by adjusting , ensuring that the system can respond quickly when skidding and reducing the risk of vehicle loss of control.

[0052] Weight coefficient value: After many experiments, the optimal , , value: : Provides greater adjustment response when the friction coefficient changes greatly.

[0053] : Reduce the severity of power output adjustments at higher speeds.

[0054] : When skidding is severe, quickly adjust the system to reduce the risk of skidding.

[0055] like Figure 2 As shown, an unmanned driving intelligent analysis system is used to implement the above-mentioned unmanned driving intelligent analysis method, including the following modules: Multi-dimensional data acquisition module, used to collect multi-dimensional data in real time through vehicle sensors, traffic information network and meteorological data, including vehicle speed, acceleration, slope, road friction coefficient, weather conditions; The short-term energy consumption prediction module builds a short-term energy consumption prediction model based on multi-dimensional data to predict energy consumption fluctuations in future driving routes; The driving mode dynamic adjustment module dynamically adjusts the driving mode, including acceleration mode, deceleration mode, coasting mode and regenerative braking mode, according to the energy consumption prediction results, the vehicle's battery status and future road conditions, to optimize the vehicle's energy efficiency and driving safety; Adaptive friction coefficient feedback module, which evaluates the change of road friction coefficient in real time during vehicle driving and dynamically adjusts the vehicle's power output and braking response according to the change of friction coefficient; The priority management module is used to set priorities among energy consumption fluctuations, road condition changes, energy efficiency optimization feedback adjustment mechanism and battery status perception to avoid driving mode adjustment conflicts and ensure the best performance and stability of the system in complex environments.

[0056] The present invention covers any substitution, modification, equivalent method and scheme made on the essence and scope of the present invention. In order to make the public have a thorough understanding of the present invention, specific details are described in detail in the following preferred embodiments of the present invention, but those skilled in the art can fully understand the present invention without the description of these details. In addition, in order to avoid unnecessary confusion about the essence of the present invention, well-known methods, processes, procedures, components and circuits are not described in detail.

[0057] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. An unmanned driving intelligent analysis method, characterized in that: The following steps are involved: S1, multi-dimensional data collection and prediction: Through on-board sensors, traffic information networks and meteorological data, multi-dimensional data including road conditions, slope, road friction coefficient and weather conditions are collected in real time, and a short-term energy consumption prediction model is built based on the multi-dimensional data. The short-term energy consumption prediction model predicts energy consumption fluctuations in future driving routes; S2, real-time driving mode dynamic adjustment: According to the energy consumption fluctuation prediction results of the S1 medium- and short-term energy consumption prediction model, and based on the multi-modal energy efficiency optimization mechanism, the driving mode of the unmanned vehicle is dynamically adjusted, including acceleration mode, deceleration mode, gliding mode, and regenerative braking mode; the dynamic adjustment not only considers the current energy consumption, but also combines the vehicle battery status, current speed and future road conditions to optimize the vehicle's driving status in real time; S3, adaptive road friction feedback and driving strategy optimization: Based on the real-time road friction coefficient collected in S1, the adaptive friction coefficient feedback mechanism dynamically evaluates the changes in the road friction coefficient between the tire and the ground under different road conditions, and adjusts the vehicle's power output and braking response to ensure optimal energy efficiency and driving safety in wet, icy or loose road conditions.

2. The unmanned driving intelligent analysis method according to claim 1, characterized in that: The multi-dimensional data collection of S1 includes: On-board sensor data collection: On-board sensors, including lidar, camera, millimeter-wave radar, and inertial measurement unit, collect road condition information and slope data of the vehicle's surrounding environment in real time; lidar is used to measure road obstacles and road morphology, and the inertial measurement unit captures the vehicle's tilt angle and acceleration changes to detect slope changes; Traffic information network data acquisition: through the connection with the external traffic information network, the current road traffic flow, congestion, accident information and road construction conditions can be obtained in real time; Weather condition collection and analysis: Through data connection with the meteorological network, the current weather conditions of the environment are obtained in real time, including temperature, wind speed and humidity parameters. The weather conditions are used to analyze the weather conditions that affect vehicle driving; Real-time estimation of road friction coefficient: Through the data analysis of the on-board wheel speed sensor and the braking system, combined with the real-time road feedback, the current road friction coefficient is inferred and cross-validated with the historical data of the traffic network to ensure the accuracy of the friction coefficient estimation. The real-time estimation of the road friction coefficient is expressed as: ,in, is the road friction coefficient estimated in real time, is the traction between the vehicle and the ground, is the vehicle mass, is the acceleration due to gravity.

3. The unmanned driving intelligent analysis method according to claim 2, characterized in that: The construction of the S1 short-term energy consumption prediction model specifically includes: The collected multi-dimensional data is used as characteristic data related to energy consumption, including road conditions, slope, road friction coefficient and weather conditions; A short-term energy consumption prediction model is constructed using a multi-layer feedforward neural network. By inputting feature data, the model is trained to predict energy consumption fluctuations in future driving routes. The short-term energy consumption prediction model is continuously iterated and updated based on historical data and real-time collected data to improve prediction accuracy. Based on the trained short-term energy consumption prediction model, real-time data of the current road section is input to predict energy consumption fluctuations in future driving routes, and energy consumption prediction results for each road section are output.

4. The unmanned driving intelligent analysis method according to claim 3 is characterized in that: The input characteristic data of the short-term energy consumption prediction model is expressed as , including the following parameters: : The vehicle at time speed at 1000 s; : The vehicle at time The acceleration at : The vehicle at time The slope angle of the road; : Road friction coefficient, at time Real-time estimation of :time Ambient temperature at :time Ambient humidity at the time; :time Wind speed at 1000 km / h; After preprocessing, the feature data is used as the input feature of the multi-layer feedforward neural network: ； The model structure of the short-term energy consumption prediction model includes: Input layer: The input layer receives Vector with dimension , that is, the number of input features, the input dimension ; Hidden layer: Each hidden layer consists of multiple neurons. hidden layers, and the number of neurons in each layer is ,in , the calculation formula for each layer is: ,in, For the The output of the hidden layer, For the The weight matrix of the layer has the dimension , represents the weight between the current layer and the previous layer of neurons, For the The bias term of the layer has the dimension , ReLU is the activation function; The number of neurons in each layer Set according to task complexity; Output layer: The output layer predicts the energy consumption value of the vehicle at the current moment , the output of this layer is calculated by a linear activation function: ,in, is the weight matrix of the output layer, with dimension , mapping the output of the hidden layer to a single energy consumption value, is the bias term of the output layer, with a dimension of 1×1, For time The predicted value of energy consumption at When training the model, minimize the energy consumption prediction error and use the mean square error as the loss function; After the model training is completed, enter the current time The eigenvector of , calculate the energy consumption value , using sliding windows to predict the future Energy consumption at each moment: ； in, It is the input feature of the future road section. By inputting the multi-dimensional data of the future, the energy consumption value at different time points is predicted. It is a multi-layer feed-forward neural network.

5. The unmanned driving intelligent analysis method according to claim 1, characterized in that: The real-time dynamic adjustment of the driving mode in S2 specifically includes: S21, based on the short-term energy consumption prediction model, real-time analysis of energy consumption fluctuations , select driving mode, consumption fluctuation calculation is based on the energy consumption forecast value at adjacent moments , driving mode selection is based on energy consumption fluctuations ,when When the energy consumption increases, consider the deceleration mode or the gliding mode. When , it means that the energy consumption is reduced and the current driving mode is maintained; S22, battery status perception and optimization adjustment: Battery status perception is achieved by combining the battery power status Analyze the relationship between battery power status and energy consumption to maintain energy efficiency balance by optimizing driving mode in real time. When the vehicle is in a state of slowing down, the glide mode or regenerative braking mode is preferred to extend the driving range. When , the acceleration mode is allowed; S23, optimization and adjustment of future road conditions and driving status: In future road conditions, including the impact of slope changes on vehicle energy consumption, the calculation formula is as follows: ,in, The vehicle is at a slope angle The power consumption under is the vehicle mass, is the acceleration due to gravity, For time The speed of the vehicle at For time The slope angle of the road; when the vehicle encounters an uphill slope, the slope Increase, power consumption The vehicle speed also increases accordingly, and the deceleration mode is selected first; when the vehicle encounters a downhill slope, the coasting mode is selected to reduce energy consumption; S24, energy efficiency optimization feedback adjustment mechanism: In the energy efficiency optimization feedback mechanism, dynamic adjustments are made according to the real-time state and energy consumption of the vehicle. The energy efficiency is calculated as follows: ,in, For time The energy efficiency is the speed per unit of energy consumption. is the instantaneous speed of the vehicle, For time By monitoring energy efficiency , real-time switching between driving modes, when When the vehicle speed is lower than the preset threshold, it indicates low energy efficiency and the vehicle will be decelerated or coasted to improve energy efficiency. When the preset threshold is exceeded, it indicates that the vehicle is energy-efficient in the current mode and maintains the current driving state.

6. The unmanned driving intelligent analysis method according to claim 5, characterized in that: The real-time driving mode dynamic adjustment further includes preset adjustment priorities, and the adjustment priorities are arranged as follows: Battery status perception and optimization adjustment > Future road conditions and driving status optimization adjustment > Energy consumption fluctuation adjustment > Energy efficiency optimization feedback adjustment mechanism.

7. The unmanned driving intelligent analysis method according to claim 2, characterized in that: The S3 specifically includes: S31, friction coefficient based on real-time acquisition The adaptive feedback mechanism dynamically evaluates the friction changes under different road conditions and calculates the rate of change of the friction coefficient. To identify rapid fluctuations in friction, it is calculated as: ,according to The size and trend of the vehicle can be used to determine the changes in the current road conditions and provide feedback information for the power output and braking systems; S32, based on friction coefficient and its rate of change , adjust vehicle power output in real time When the friction coefficient decreases, reduce the vehicle's power output to reduce the risk of wheel slippage; when the friction is restored, gradually increase the power output to restore normal driving conditions. S33, based on real-time friction coefficient Dynamically adjust the vehicle's braking response When the friction force decreases, the feedback response of the braking system is increased to avoid excessive braking force causing wheel lock or loss of control; when the friction coefficient increases, the braking feedback is reduced to ensure the smoothness of the braking operation. The formula for adjusting the braking response is: , ensuring that the braking response is more sensitive when the friction coefficient is low, preventing excessive braking force; when the friction coefficient is high, the system allows greater braking force.

8. The unmanned driving intelligent analysis method according to claim 7, characterized in that: The vehicle power output adjustment in S32 is expressed as: ,in, is the adjustment factor, is the current road friction coefficient. As the friction coefficient decreases, the power output decreases proportionally, and vice versa, it gradually returns to normal.

9. The unmanned driving intelligent analysis method according to claim 8, characterized in that: The adjustment factor According to the friction coefficient of the road The change range of the vehicle's current speed , and wheel slip to dynamically adjust the adjustment coefficient Calculated as: ,in, is the rate of change of friction coefficient, is the current speed of the vehicle, The wheel slip detection parameter reflects the wheel slip condition through the slip degree of the wheel contacting the ground. When it increases, the adjustment coefficient should increase. , , It is an empirical weight coefficient determined through experimental data to balance the effects of changes in friction coefficient, speed, and slippage on the adjustment range.

10. An unmanned driving intelligent analysis system, used to implement an unmanned driving intelligent analysis method as claimed in any one of claims 1 to 9, characterized in that: Includes the following modules: Multi-dimensional data acquisition module, used to collect multi-dimensional data in real time through vehicle sensors, traffic information network and meteorological data, including vehicle speed, acceleration, slope, road friction coefficient, weather conditions; A short-term energy consumption prediction module, which constructs a short-term energy consumption prediction model based on the multi-dimensional data to predict energy consumption fluctuations in future driving routes; A driving mode dynamic adjustment module dynamically adjusts the driving mode, including acceleration mode, deceleration mode, gliding mode and regenerative braking mode, according to the energy consumption prediction result, the battery status of the vehicle and the future road conditions, so as to optimize the energy efficiency and driving safety of the vehicle; Adaptive friction coefficient feedback module, which evaluates the change of road friction coefficient in real time during vehicle driving and dynamically adjusts the vehicle's power output and braking response according to the change of friction coefficient; The priority management module is used to set priorities among energy consumption fluctuations, road condition changes, energy efficiency optimization feedback adjustment mechanism and battery status perception to avoid driving mode adjustment conflicts.

Images