PCC-based fuel-saving auxiliary driving system and method

Abstract

The invention discloses an oil-saving auxiliary driving system and method based on PCC, and belongs to the technical field of auxiliary driving, and the system comprises a data collection module which is used for obtaining the driving data, environment data and navigation data of a vehicle in real time; the background calculation module is in communication connection with the data acquisition module and is used for processing the acquired data; the user interaction module is in communication connection with the background calculation module and is used for presenting the fuel-saving driving suggestion and the fuel consumption evaluation information; and the vehicle economic driving scoring module calculates the economic score of each time of driving according to the actual fuel consumption condition and the suggested value of the vehicle, and performs ranking and reward in a certain period, so that the limitation problem that only single driving behavior data is concerned in the traditional technology is effectively solved, and meanwhile, the fuel-saving effect is improved.

Description

Technical Field

[0001] This invention belongs to the field of driver assistance technology, specifically relating to a fuel-saving driver assistance system and method based on PCC. Background Technology

[0002] With the improvement of people's living standards and the acceleration of urbanization, vehicles have become an important tool for daily travel. However, with the finite and non-renewable nature of petroleum resources and the continuous rise in fuel prices, high-fuel-consumption driving habits are becoming increasingly prominent, seriously affecting energy conservation, environmental protection, and user economy. According to statistics, fuel consumption costs account for approximately 47% of the total cost of medium and heavy-duty commercial trucks throughout their entire life cycle. If a tractor-trailer with an annual operating mileage of 150,000 kilometers can reduce fuel consumption by 1 liter per 100 kilometers, it can save about 9,000 yuan per year. This not only increases the user's vehicle operating costs but also has an adverse impact on the environment, contradicting the national energy conservation and emission reduction policies.

[0003] Currently, while some fuel-saving driving analysis methods exist in the market, they have many shortcomings. Traditional technologies typically focus only on single driving behavior data, such as driving speed or braking frequency, failing to comprehensively and deeply explore the driver's driving style and behavioral characteristics. This results in route planning methods that cannot provide drivers with route recommendations that truly match their driving habits. Furthermore, existing driving behavior analysis methods are often based on single road conditions or fixed load conditions, failing to accurately describe fuel consumption during driving, leading to poor fuel-saving effects. Simultaneously, the lack of full utilization of real-time traffic information means that current road congestion, traffic accidents, and other real-time conditions cannot be reflected in a timely manner, affecting the effectiveness of route optimization. Summary of the Invention

[0004] To address the aforementioned issues, the current product development incorporates a vehicle fuel-saving system that integrates navigation information and big data technology. This system divides map segments into 50-kilometer intervals, defines macro-level driving profiles based on historical vehicle driving records, and uses the average value of vehicles with below-average fuel consumption to generate suggested driving targets. Simultaneously, based on navigation information, it provides driving alerts and suggestions regarding traffic congestion, weather conditions, and road changes ahead, helping drivers choose the most economical driving maneuvers. This approach not only improves fuel efficiency but also encourages good driving behavior and provides correction and optimization suggestions for poor driving habits, thereby achieving true energy conservation and emission reduction.

[0005] The technical solution of the present invention is as follows: A fuel-saving assisted driving system based on PCC, comprising:

[0006] The data acquisition module is used to acquire vehicle driving data, environmental data, and navigation data in real time.

[0007] The background computing module is communicatively connected to the data acquisition module to process the acquired data;

[0008] The user interaction module is communicatively connected to the backend computing module and is used to present the fuel-saving driving suggestions and fuel consumption assessment information.

[0009] The vehicle economy driving score module calculates the economy score for each drive based on the vehicle's actual fuel consumption and recommended values, and ranks and rewards drivers over a certain period.

[0010] As an improvement of the present invention, the background calculation module is also used to calculate the driver's economic driving score based on the deviation value and the preset weight coefficient.

[0011] The system also includes a scoring and incentive module for providing feedback and incentives to the driver based on the economic driving score.

[0012] As an improvement of the present invention, the calculation model for the economic driving score is as follows:

[0013] (1),

[0014] Where f(x) is the economic driving score, For factors, The corresponding weights of each factor.

[0015] As an improvement of the present invention, the training process of the multiple linear regression model includes: constructing a dataset, wherein the feature variables include standardized driving behavior feature values ​​and the target variable is the standardized actual fuel consumption value; and solving for the optimal weight coefficient vector of each driving behavior feature by minimizing the error between the predicted value and the true value.

[0016] As an improvement of the present invention, a fuel-saving driving assistance method implemented by the system includes the following steps:

[0017] Step 1: Acquire real-time vehicle driving behavior data, environmental data, and navigation data;

[0018] Step two: Calculate the recommended fuel-saving value for each route;

[0019] Step 3: Based on the vehicle's current driving behavior and environmental factors, provide real-time driving suggestions to the vehicle;

[0020] Step four: After a certain amount of driving distance has been covered, an economic driving behavior score is generated.

[0021] As an improvement to the present invention, it also includes:

[0022] After completing the route unit, an economic driving score is calculated based on the deviation value and the weighting coefficients determined by the multiple linear regression model.

[0023] Based on the rating, a driving behavior summary report and optimization suggestions are generated.

[0024] As an improvement of the present invention, the weighting coefficient is determined through the following steps:

[0025] Collect driving data samples from historical mileage units, extract driving behavior features as feature variables, and obtain the corresponding fuel consumption as the target variable;

[0026] The dataset was trained using a multiple linear regression algorithm to obtain weight coefficients that reflect the degree of influence of each driving behavior feature on fuel consumption.

[0027] As an improvement of the present invention, the reference feature vector is obtained by calculating the average driving behavior characteristics of a group of vehicles with fuel consumption below the average value on the same road unit.

[0028] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0029] 1. By dividing the map road segment into 50-kilometer segments and defining a macro-route profile target for fuel-saving driving based on historical vehicle driving records, this invention can comprehensively and deeply explore the driver's driving style and behavioral characteristics in specific road environments through big data technology. This enables more accurate driving behavior suggestions during driving in specific road environments, effectively solving the limitation of traditional technologies that only focus on single driving behavior data.

[0030] 2. This invention forms suggested driving targets by averaging the average fuel consumption of vehicles with average fuel consumption values ​​below a specific route, and provides driving reminders and suggestions based on navigation information regarding route congestion, weather effects, and changes in the road ahead. This overcomes the limitations of existing technologies in analyzing road conditions or fixed load conditions without relying on road conditions, and improves fuel-saving performance.

[0031] 3. This invention utilizes big data to collect and analyze vehicle driving status, environmental data, and navigation data in real time, which can promptly reflect the current real-time conditions such as road congestion and traffic accidents. It effectively solves the problem that existing economic driving technologies lack full utilization of real-time traffic information and significantly improves the effect of route optimization.

[0032] 4. This invention encourages drivers’ excellent driving behavior based on quantitative scoring, and provides correction and optimization suggestions for drivers’ poor driving behavior, thereby realizing comprehensive monitoring and guidance of driving behavior and effectively solving the shortcomings of existing technologies in providing effective encouragement and correction of driving behavior.

[0033] 5. This invention calculates fuel-saving strategy rules in real time and combines them with the vehicle's smart cockpit or APP mini-program to provide drivers with PCC assisted driving reminders. This provides drivers with comprehensive driving information and optimization suggestions, overcoming the shortcomings of existing technologies in terms of insufficient accuracy of fuel consumption prediction models, and achieving more accurate fuel-saving guidance. Attached Figure Description

[0034] Figure 1 This is the output process for the historical driving guidance report of this invention;

[0035] Figure 2 The output process for real-time driving behavior optimization in this invention is as follows. Detailed Implementation

[0036] The present invention will be further illustrated below with reference to the accompanying drawings and specific embodiments. It should be understood that the following specific embodiments are for illustrative purposes only and are not intended to limit the scope of the invention.

[0037] Example:

[0038] This embodiment is a fuel-saving driving assistance system and method based on PCC. The system includes a data acquisition module, a background calculation module, a user interaction module, and a scoring module.

[0039] The data acquisition module acquires real-time vehicle driving behavior data, environmental data, and navigation data, including vehicle speed, fuel consumption, braking frequency, acceleration frequency, lane changing frequency, as well as information such as whether there is a traffic jam and weather conditions.

[0040] In this embodiment, the background calculation module processes the collected data, specifically including the following steps:

[0041] 1) Divide the map road segments into 50-kilometer intervals to create a route profile database;

[0042] 2) Based on historical vehicle driving records, define macro-target parameters for each route profile, including average speed, number of rapid accelerations, number of rapid decelerations, number of lane changes, etc.

[0043] 3) Extract features from the actual driving data for each route and construct a feature vector;

[0044] 4) Based on the vehicle's fuel consumption, group the feature vectors and calculate the average value of each group of feature vectors as reference data;

[0045] 5) Compare the current vehicle's feature vector with the reference data and calculate the deviation value of the fuel consumption influencing parameters;

[0046] 6) Based on factors such as deviation value, road condition information, and quality information, calculate the recommended fuel-saving value for each route.

[0047] The user interaction module includes the vehicle's smart cockpit and a mobile application, which, in conjunction with the navigation function, displays the following information related to fuel-efficient driving:

[0048] Real-time driving guidance in the example:

[0049] Based on the vehicle's current driving behavior and environmental factors, provide real-time driving suggestions:

[0050] 1) Display the real-time fuel consumption of vehicles on the current road segment, and compare the real-time fuel consumption with the profile data of the route to quantitatively assess whether the current fuel consumption is economical;

[0051] 2) Based on road environment factors, such as whether there are rugged mountain roads ahead, whether there is a highway entrance or exit, whether there is a turn, whether there are uphill or downhill sections, etc., provide guidance for current driving behavior, such as suggesting to slow down in advance, suggesting to accelerate appropriately, suggesting to switch to the right lane, suggesting to switch to the left lane, and suggesting not to change lanes frequently, etc.

[0052] Summary of historical driving behavior:

[0053] After a certain amount of driving distance has been covered, the historical driving behavior that occurred during this distance is summarized and quantitatively evaluated based on the vehicle driving profile. An economic driving behavior score is generated, and the driver's uneconomical driving behavior is summarized to provide driving optimization suggestions and suggestions for improving driving habits.

[0054] Fuel-saving tips for a 50km journey: After driving 50km, profile the data of the current 50km journey to determine the average fuel consumption level of all similar vehicles traveling on this route.

[0055] 2) Daily fuel-saving suggestions: After completing the daily driving (the vehicle's daily driving distance is >100 kilometers), based on the vehicle's driving profile for the day, determine whether the fuel-saving performance was excellent, and whether there were any uneconomical driving behaviors that deserve attention, such as abnormal long-term low-speed driving.

[0056] Description of relevant algorithm model modules:

[0057] I. Explanation of the use of models in immediate guidance:

[0058] In real-time guidance, vehicle alerts are primarily defined by various rule models based on historical data analysis, such as:

[0059] When the vehicle speed is greater than 70 km / h and the distance to a road change point such as a turn or exit from the highway is less than 1 km, remind the vehicle to slow down appropriately.

[0060] When there is no congestion 1km ahead, the road conditions will not change significantly 3km ahead, the vehicle is traveling on the highway, and the average speed over the past 10 minutes is below 50km / h, and the current speed is below 60km / h, remind the vehicle to accelerate appropriately.

[0061] When a rugged mountain road appears 1km ahead and the vehicle speed is >70km / h, please remind the driver to slow down appropriately.

[0062] II. Explanation of the algorithm model used in historical driving guidance:

[0063] The vehicle economy driving score module calculates an economy score for each drive based on the vehicle's actual fuel consumption and recommended values, and ranks and rewards drivers over a certain period. Specifically, it includes:

[0064] 1) Divide each driving session into 50-kilometer driving segments;

[0065] 2) Extract feature vectors and design models for each 50-kilometer driving segment cycle;

[0066] 3) Compare the feature vectors with the reference data and calculate the deviation score of the fuel consumption influencing parameters;

[0067] 4) Based on factors such as deviation score and road conditions, and with appropriate weights, calculate the economic driving score for each road segment. The calculation formula is as follows:

[0068] (1),

[0069] f(x) is the economic driving score;

[0070] These are factors such as various partial deviations and road conditions; The corresponding weights of each factor;

[0071] The weights in the calculation formula are derived from a dataset of historical vehicle driving data for the current road segment. A linear regression algorithm is used to design the model, and the weight factors are obtained as follows:

[0072] 1. Dataset construction.

[0073] 1) Data collection: Through OBD, AEB, and T-box in-vehicle equipment, a large number of vehicles' complete driving records on the same or similar road sections that have been divided into 50-kilometer segments are collected.

[0074] Sample definition: A 50-kilometer journey by a vehicle is used as a data sample for that road segment.

[0075] Feature variable (X): The driving behavior feature vector extracted for each sample, including but not limited to:

[0076] X1: Average vehicle speed; x2: Number of rapid accelerations (acceleration > 2 m / s²); x3: Number of rapid decelerations (deceleration < -2 m / s²); x4: Number of lane changes; x5: Idle time percentage; x6: Vehicle type: tractor truck, van, passenger car, and more features can be added as needed.

[0077] Target variable y: the actual fuel consumption s per unit mileage corresponding to this sample, and the fuel consumption s is standardized using Z-score transformation. ,

[0078] After standardization, to facilitate understanding and interpretation, we further convert the Z-score to a T-score. We design a standard score T-score with a mean of 50 and a standard deviation of 10: T-score = 50 + 10 * Z-score.

[0079] The T-score, representing the final value of the target variable y, is the target that the model aims to predict.

[0080] The final structured dataset is: D={(X_i,y_i)|i=1,2,...,N}, where N is the total number of samples.

[0081] 2. Model design: Multiple linear regression.

[0082] Establish the following linear regression model: y_pred=w0+w1·x1+w2·x2+w3·x3+...+w·x; where:

[0083] y_pred is the fuel consumption per unit mileage predicted by the model;

[0084] w0 is the intercept term (baseline fuel consumption, reflecting inherent factors such as vehicle base fuel consumption and road conditions);

[0085] w1, w2, ..., w are the regression coefficients to be solved, i.e. the weighting factors of each driving behavior feature;

[0086] x1, x2, ..., x are the standardized driving behavior feature values. The feature values ​​are first standardized by Z-score of the original variable fields to make the features comparable.

[0087] 3. Model training and weight calculation.

[0088] Loss function: The mean squared error (MSE) is used as the loss function: Loss=(1 / N)Σ(y_true-y_pred)².

[0089] Optimization algorithm: Use least squares (OLS) or gradient descent to minimize the loss function and solve for the optimal weight vector W'=[w1',w2',...,w'].

[0090] Regularization (optional): To prevent overfitting, L2 regularization (Ridge regression) can be introduced: Loss = MSE + λΣwᵢ²,

[0091] Where λ is the regularization intensity hyperparameter.

[0092] 4. Explanation and application of weighting factors.

[0093] Symbol explanation:

[0094] If wᵢ'>0: This indicates that the driving behavior is positively correlated with fuel consumption; the more frequent the behavior, the higher the fuel consumption.

[0095] If wᵢ'<0: This is theoretically unreasonable and may indicate a problem with feature engineering or strong multicollinearity, which needs to be investigated.

[0096] Numerical magnitude: The larger the absolute value of |wᵢ'|, the more significant the impact of the driving behavior on fuel consumption.

[0097] Practical applications:

[0098] After training, the obtained w1', w2', ..., w' are the scientific weighting factors for each driving behavior feature.

[0099] When calculating the deviation score, these regression coefficients are directly used as weight coefficients _i:

[0100] The single-feature deviation score _i = |standardized value _i| × |wᵢ'|; taking the absolute value ensures that the deviation score is always positive.

[0101] 5. Model validation and updates.

[0102] Verification method:

[0103] The dataset was divided into a training set (70%) and a test set (30%).

[0104] Calculate metrics such as R² coefficient of determination and MAE (mean absolute error) on the test set to evaluate the model's prediction accuracy.

[0105] Dynamic automatic updates:

[0106] With the continuous accumulation of new data due to seasonal changes and the addition of new car models, the model is retrained and the weight factors are updated regularly every quarter to ensure its timeliness and accuracy.

[0107] Scene-by-scene modeling (advanced):

[0108] Based on the actual results of data collection, independent regression models can be constructed for different road types and weather conditions to obtain more refined weighting factors.

[0109] Drivers whose scores exceed a set threshold will receive a reward notification; for drivers whose scores exceed a set threshold, targeted driving optimization suggestions will be proposed based on single-factor analysis rules in the rule base.

[0110] It should be noted that the above content merely illustrates the technical concept of the present invention and should not be construed as limiting the scope of protection of the present invention. For those skilled in the art, various improvements and modifications can be made without departing from the principle of the present invention, and all such improvements and modifications fall within the scope of protection of the claims of the present invention.

Claims

1. A PCC-based fuel-saving assistant driving system, characterized by, The system comprises: a data acquisition module for acquiring driving data, environmental data and navigation data of the vehicle in real time; a background calculation module in communication connection with the data acquisition module for processing the collected data; a user interaction module in communication connection with the background calculation module for presenting the fuel-saving driving suggestions and fuel consumption evaluation information; a vehicle economic driving score module for calculating the economic driving score of each driving based on the actual fuel consumption and the recommended value, and ranking and rewarding within a certain period.

2. The system of claim 1, wherein, The background calculation module is further configured to calculate the economic driving score of the driver based on the deviation value and a preset weight coefficient. The system further comprises a score incentive module for feeding back and motivating the driver according to the economic driving score.

3. The system of claim 2, wherein, The calculation model of the economic driving score is: （1）， where f(x) is an economic driving score, is a factor, is a corresponding weight for each factor.

4. The system of claim 3, wherein, The training process of the multiple linear regression model comprises: constructing a data set, wherein the characteristic variables include the standardized driving behavior characteristic values, and the target variable is the standardized actual fuel consumption value; and solving the optimal weight coefficient vector of each driving behavior characteristic by minimizing the error between the predicted value and the true value.

5. A PCC-based fuel-saving assisted driving method implemented by the system of any one of claims 1 to 4, characterized in that, The method comprises the following steps: Step 1: acquiring vehicle driving behavior data, environmental data and navigation data in real time; Step 2: calculating the fuel-saving recommendation value for each trip; Step 3: giving real-time driving suggestions based on the current driving behavior and environmental factors of the vehicle; Step 4: forming an economic driving behavior score after a certain amount of driving.

6. The method of claim 5, wherein, Further comprising: After completing the trip unit, calculating the economic driving score based on the deviation value and the weight coefficient determined by the multiple linear regression model; Generating a driving behavior summary report and optimization suggestions according to the score.

7. The method of claim 6, wherein, The weight coefficient is determined by the following steps: Collecting driving data samples of historical trip units, extracting driving behavior characteristics as characteristic variables, and obtaining corresponding fuel consumption as target variables; Training the data set using a multiple linear regression algorithm to obtain weight coefficients reflecting the influence of each driving behavior characteristic on fuel consumption.

8. The method of claim 7, wherein, The reference feature vector is obtained by calculating the average driving behavior characteristics of the vehicle group with fuel consumption lower than the average value in the same trip unit.

Images