Evaluation and prediction method of road friction coefficient

Abstract

The present invention discloses a road friction coefficient evaluation and prediction method, comprising the following steps: S1, collecting vehicle and environment information; S2, analyzing and evaluating basic friction coefficient data; S3, measuring road friction coefficient; S4, evaluating friction coefficient big data; S5, predicting friction coefficient warning. The present invention can obtain lane-level real-time road friction coefficient evaluation values ​​accurate to complex environments, which is closer to the actual situation, and constructs a road friction coefficient Internet of Vehicles system including an on-board terminal, a big data cloud computing processing platform, and a road friction coefficient warning platform, which can be used for real-time road friction coefficient warnings for road vehicles and provide medium- and long-term solutions.

Description

Technical Field

[0001] The invention belongs to the technical field of safe driving in urban traffic road engineering, and in particular relates to a method for evaluating and predicting a road friction coefficient. Background Art

[0002] Due to factors such as vehicle load and environment, the anti-skid performance between vehicles and roads is weakened, often leading to frequent traffic accidents. According to research surveys in the UK, Japan and other countries, traffic accidents caused by the decline in anti-skid performance of roads account for about 1 / 5 of the total number of accidents each year. With the rapid development of my country's national economy, the traffic volume and speed of high-grade highways and urban roads in my country are large, and the requirements for the anti-skid ability of roads have increased accordingly. Highway construction and maintenance management departments have paid more and more attention to the impact of the anti-skid performance of roads on traffic safety.

[0003] At present, my country faces many problems in the testing and evaluation of the anti-skid performance of pavement: First, the sand-laying method, pendulum friction instrument method, and DF tester method in the structural depth method use manual methods for detection, which have low detection efficiency, and the structural depth is only an indirect indicator. The laser scanning structural depth method and the lateral force coefficient tester method have high automation and high detection efficiency, but the test is easily affected by the external environment, and the test results are easily affected by the test road environment such as water, snow, mud, etc. In addition, it is also necessary to test in closed traffic or low traffic volume. The established distance method, locked wheel trailer method and other test methods are mainly used for the friction between the vehicle tires and the road surface after the vehicle is braked and the wheels are locked. The automation level and efficiency are high, but they can only be carried out in the form of annual inspections, with a long window period, the working conditions are public welfare, and the real-time performance is insufficient. The road friction force under the vehicle braking state is directly related to driving safety. At this time, the friction between the tire and the road surface is different from the traditional static friction and dynamic friction. It is creeping friction and is directly related to the vehicle speed. Therefore, testing the friction coefficient of the vehicle under the actual braking state is of great significance to the anti-skid ability of the road surface.

[0004] The rapid development of Internet of Vehicles technology and big data has enabled the information of vehicles and roads to be perceived in real time, and to be aggregated, analyzed and used. Since the current domestic road anti-skid performance detection is still in the traditional manual or automated stage, the anti-skid friction coefficient data provided is lagging, intermittent and one-sided, while the actual vehicle and road surface status is real-time. Traditional anti-skid data cannot provide reliable warning of insufficient anti-skid ability in braking state for vehicle driving. Therefore, a road friction coefficient evaluation and early warning prediction method based on vehicle braking state and Internet of Vehicles big data is proposed. A road friction coefficient estimation model is constructed based on big data analysis of massive data to realize real-time estimation of lane-level road friction coefficient in complex environments. On the one hand, the big data analysis platform library and road friction coefficient estimation model are continuously updated through historical data. On the other hand, the instantaneous road friction coefficient estimation value and threshold are provided according to the actual driving environment. Timely warning and solutions are provided for the decline of road friction coefficient to avoid traffic safety accidents caused by the decline of road anti-skid ability. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide a method for evaluating and predicting the road friction coefficient. The road friction coefficient estimation model constructed based on big data analysis can evaluate the road friction coefficient of vehicles in complex environments in real time, and provide early warning information and solutions for insufficient road friction for moving vehicles.

[0006] A method for evaluating and predicting a road friction coefficient comprises the following steps:

[0007] S1. Collect vehicle and environment information;

[0008] S2. Analyze and evaluate basic data of friction coefficient;

[0009] S3, calculating the road friction coefficient;

[0010] S4, big data evaluation of friction coefficient;

[0011] S5. Early warning prediction of friction coefficient.

[0012] Furthermore, in S1, the vehicle and environment information are sensed and collected, and the data information of the vehicle during the braking state is monitored through the vehicle monitoring system and various sensors to obtain relevant data of the vehicle and the road during the braking process.

[0013] Furthermore, in S2, the basic data of the friction coefficient collected in S1 is located and uploaded to the cloud platform, and the data collected by vehicle perception at different locations is uploaded to the big data cloud platform through the Internet of Vehicles system, stored by category, and further processed, analyzed and fed back.

[0014] Furthermore, in S3, the road friction coefficient formula under the vehicle braking state is as follows:

[0015]

[0016] a=a1+a2

[0017]

[0018] Where, μ is the estimated road friction coefficient,

[0019] μ1——Friction coefficient between brake pad and brake disc,

[0020] F N ——Braking force, in N,

[0021] r——brake wheel radius, in m,

[0022] r1——brake pad radius, in m,

[0023] m——brake wheel mass, in kg,

[0024] w——vehicle load, in N,

[0025] a——tire rotation deceleration, in rad·s -2 ,

[0026] a1——Tire rotation deceleration caused by wind resistance, in rad·s -2 ,

[0027] a2——Tire rotation deceleration caused by braking, in rad·s -2 ,

[0028] f——wind resistance, unit is N.

[0029] Furthermore, in S4, the friction coefficient big data assessment collects the estimated values ​​of the road friction coefficient under different vehicles, speeds, positions, humidity and temperatures, and calculates the average friction coefficient under different conditions; through the accumulation of massive data, a real-time assessment model for the road friction coefficient is constructed, and the road friction coefficient thresholds under different conditions are designed.

[0030] Furthermore, in S5, the friction coefficient warning prediction gives a real-time vehicle friction coefficient threshold value based on the real-time vehicle and environmental conditions of the trip. When the estimated road friction coefficient during vehicle driving exceeds the threshold, the big data cloud platform will remotely send a warning signal to the vehicle through a wireless sending module and provide corresponding solutions (instantaneous deceleration, mid-term replacement of tires or brake pads, long-term replacement of road surface).

[0031] The beneficial effects of the present invention are:

[0032] (1) The present invention monitors information related to vehicle-road friction and uses a wireless transmission module to collect the collected data to a big data platform;

[0033] (2) The road friction coefficient evaluation method provided by the present invention is based on the data volume of the big data platform. After deep processing, a road friction coefficient estimation model under complex conditions is established, and thresholds under different conditions are given;

[0034] (3) The present invention provides a road friction coefficient early warning prediction method that gives a real-time estimated value and threshold of the road friction coefficient, provides early warning prompts and deceleration suggestions, and provides drivers and road maintenance departments with medium- and long-term solutions based on long-term accumulated data;

[0035] (4) The present invention constructs a road friction coefficient evaluation and early warning prediction method based on vehicle braking status and Internet of Vehicles big data, which can obtain road friction coefficient evaluation values ​​accurate to the lane level, in real time and in complex environments, which is closer to the actual situation. It also constructs an Internet of Vehicles system for road friction coefficient evaluation and early warning, which can be used for real-time road friction coefficient monitoring and early warning of road vehicles and provide medium- and long-term solutions. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 This is a working diagram of the vehicle networking system for road friction coefficient evaluation and early warning of the present invention;

[0037] Figure 2 This is a schematic diagram of the layout of the vehicle-mounted terminal of the present invention;

[0038] Figure 3 It is a schematic diagram of the calculation parameters of the road friction coefficient of the present invention;

[0039] Figure 4 It is a schematic diagram of the big data cloud computing processing platform of the present invention;

[0040] Figure 5 A schematic diagram of a road friction coefficient early warning platform of the present invention;

[0041] Figure 6 It is a hot spot distribution map of the lane-level road surface friction coefficient of the present invention;

[0042] Figures 1 to 6In: 1. On-board terminal, 1.1. Vehicle information, 1.2. Vehicle sensor, 1.3. Environmental information, 1.4. Environmental sensor, 1.5. Road monitoring equipment, 1.6. Road information, 2. Mobile base station, 3. Wireless data communication gateway, 4. Internet, 5. Acquisition and control gateway, 6. Big data cloud computing processing platform, 7. Road friction coefficient early warning platform, 7.1. Road friction coefficient real-time monitoring module, 7.2. Road friction coefficient early warning prompt module, 7.3. Road friction coefficient solution supply module, 8. Lane-level road friction coefficient hot spot distribution, 8.1. Right lane one, 8.2. Right lane two, 8.3. Right overtaking lane, 8.4. Central dividing strip, 8.5. Left overtaking lane, 8.6. Left lane two, 8.7. Left lane one. DETAILED DESCRIPTION

[0043] like Figure 1-6 As shown, the vehicle networking system for road friction coefficient evaluation and early warning prediction scheme in the present invention includes an information-sensing vehicle terminal 1, a mobile base station 2 in a wireless transmission module, a wireless data communication gateway 3, the Internet 4 and a collection and control gateway 5, a big data cloud computing processing platform 6, and a road friction coefficient early warning platform 7. The vehicle terminal 1 collects data obtained by sensors (1.2, 1.4) and monitors 1.5, and then uses wireless transmission modules (2, 3, 4, 5) to transmit data to the big data cloud computing processing platform 6 for real-time road friction coefficient estimation and system construction under complex environments. Finally, the road friction coefficient early warning platform 7 provides early warning prompts and solutions for the vehicle terminal, so as to avoid vehicle traffic safety problems caused by insufficient road anti-skid ability.

[0044] In the present invention, the vehicle terminal 1 utilizes vehicle sensors 1.2, environmental sensors 1.4 and road monitoring equipment 1.5 to respectively collect vehicle information 1.1, environmental information 1.3 and road information 1.6, which are then aggregated to the vehicle terminal 1, transmitted to the Internet 4 through the mobile base station 2 and the wireless data communication gateway 3, and then collected to the big data cloud computing processing platform 6 for storage and analysis using the acquisition and control gateway 5.

[0045] The vehicle information 1.1 collected by the vehicle sensor 1.2 in the present invention specifically includes the vehicle load, braking force, tire rotation deceleration caused by braking, brake wheel radius, friction coefficient between brake pad and brake disc, brake pad radius, and brake wheel mass. Among them, the vehicle load, braking force, and tire rotation deceleration caused by braking are variable information that require real-time monitoring by the vehicle sensor 1.2. The brake wheel radius, friction coefficient between brake pad and brake disc, brake pad radius, and brake wheel mass are long-term constants that are acquired through the vehicle's own system.

[0046] In the present invention, the environmental information 1.3 collected by the environmental sensor 1.4 is specifically humidity, temperature and wind resistance, and the road information 1.6 collected by the road monitoring device 1.5 is specifically the real-time road name, road section location and lane of the moving vehicle.

[0047] The big data cloud computing processing platform 6 in the present invention is divided into data collection, analysis and feedback modules. The data collection module reconfigures the data collected by the vehicle terminal 1 and stores them by category, that is, classifies and stores them according to vehicle type, driving position, vehicle speed, environment (temperature + humidity) and other categories.

[0048] The data analysis module of the big data cloud computing processing platform 6 in the present invention obtains the estimated road friction coefficient by calculating the vehicle information 1.1, the environmental information 1.3 and the road information 1.6, determines the threshold value of the road friction coefficient under different conditions based on the road friction coefficient estimation system constructed according to long-term big data statistics, and continuously optimizes and updates the road friction coefficient evaluation system based on the real-time transmitted data.

[0049] The data feedback module of the big data cloud computing processing platform 6 in the present invention transmits the estimated value of the road friction coefficient to the vehicle terminal in real time according to the real-time vehicle information 1.1, environmental information 1.3 and road information 1.6 and the road friction coefficient evaluation system, and provides a reference threshold.

[0050] The estimation model construction content adopted by the real-time evaluation method of road friction coefficient in the present invention includes:

[0051] a. A training sample database for estimating road friction coefficient based on different driving conditions. Each sample data contains five components: vehicle type, speed, temperature, humidity, and vehicle position;

[0052] b. Establish a BP neural network for estimating the road friction coefficient for each driving state and perform network initialization, which includes the maximum number of training times, learning accuracy, number of hidden nodes, initial weights, thresholds, and initial learning rate;

[0053] c. Input training samples X1, X2, ..., X k , X k =[x k1 , x k2 , x k3 , x k4 , x k5 ], where k represents the number of training samples, and the five components correspond to vehicle type, vehicle speed, temperature, humidity, and vehicle position, respectively, and the training samples are evenly distributed to each Map node;

[0054] d. Use the Map function to read the network weight records saved in the Hadoop distributed file system, obtain the weight of each Map node and output it, and use the Reduce function to read the network weight records saved in the Hadoop distributed file system and receive the weight output by Map, and determine whether the next cycle is needed based on the difference between the updated network weight of the Map node and the Hadoop network weight;

[0055] e. If the network weights are no longer updated, the road friction coefficient estimation BP neural network algorithm training is completed;

[0056] f. Use the trained road friction coefficient estimation BP neural network algorithm to estimate the road friction coefficient of the vehicle terminal.

[0057] The road friction coefficient early warning platform of the present invention comprises a road friction coefficient real-time monitoring module 7.1, a road friction coefficient early warning prompt module 7.2 and a road friction coefficient solution supply module 7.3.

[0058] The road friction coefficient real-time monitoring module 7.1 of the present invention displays the real-time road friction coefficient evaluation value and threshold, including parameters such as temperature, humidity, and vehicle speed in real time, and compares the vehicle road friction coefficient value obtained by the Internet of Vehicles transmission with the threshold provided by big data.

[0059] The road friction coefficient early warning prompt module 7.2 of the present invention provides a real-time early warning prompt for the vehicle terminal, that is, a warning prompt is issued when the road friction coefficient is insufficient.

[0060] The road friction coefficient solution supply module 7.3 of the present invention provides solutions according to the early warning prompt, including an instantaneous deceleration solution, a mid-term tire and brake pad replacement solution, and a long-term road renewal solution such as an ultra-thin wear layer. The deceleration solution is a solution given immediately after the early warning information appears, the mid-term tire and brake pad replacement solution is a solution adopted when the instantaneous solution does not work well, and the long-term road renewal solution is a solution adopted when the first two solutions cannot effectively solve the problem, and the costs of the three solutions increase successively.

[0061] The present invention designs four levels of thresholds for road friction coefficients in different states (vehicle type, vehicle speed, temperature + humidity), and realizes real-time display of lane-level road friction coefficients. Figure 6 The middle shows the real-time distribution of friction coefficient of each lane road surface.

[0062] The above embodiments are only for illustrating the technical idea of ​​the present invention, and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution in accordance with the technical idea proposed by the present invention shall fall within the protection scope of the present invention.

Claims

1. A road friction coefficient evaluation and prediction method, characterized in that: The steps include: S1. Collect vehicle and environment information; S2. Analyze and evaluate basic data of friction coefficient; S3, calculating the friction coefficient of the road surface; S4, big data evaluation of friction coefficient; S5, early warning prediction of friction coefficient; Among them, in step S3, the road friction coefficient formula under the vehicle braking state is as follows: a=a1+a2 Where, μ is the estimated road friction coefficient, μ1——Friction coefficient between brake pad and brake disc, F N ——Braking force, in N, r——brake wheel radius, in m, r1——brake pad radius, in m, m——brake wheel mass, in kg, w——vehicle load, in N, a——tire rotation deceleration, in rad·s -2 , a1——Tire rotation deceleration caused by wind resistance, in rad·s -2 , a2——Tire rotation deceleration caused by braking, in rad·s -2 , f——wind resistance, in N; In step S4, the friction coefficient big data is evaluated by collecting the estimated values ​​of the road friction coefficient under different vehicles, speeds, positions, humidity and temperatures, and calculating the average values ​​of the friction coefficient under different conditions; through the accumulation of massive data, a real-time evaluation model of the road friction coefficient is constructed, and the road friction coefficient thresholds under different conditions are designed.

2. The road friction coefficient evaluation and prediction method according to claim 1, characterized in that: In S1, vehicle and environmental information perception and collection are carried out, and data information of the vehicle during braking is monitored through the vehicle monitoring system and various sensors to obtain relevant data of the vehicle and the road during braking.

3. The road friction coefficient evaluation and prediction method according to claim 1, characterized in that: In S2, the basic data of friction coefficient collected in S1 is located and uploaded to the cloud platform. The data collected by vehicle perception at different locations are uploaded to the big data cloud platform through the Internet of Vehicles system, stored by category, and further processed, analyzed and fed back.

4. The method for evaluating and predicting road friction coefficient according to claim 1, characterized in that: In S5, the friction coefficient warning prediction gives a real-time vehicle friction coefficient threshold based on the real-time vehicle and environmental conditions of the trip. When the estimated road friction coefficient during vehicle driving exceeds the threshold, the big data cloud platform will remotely send a warning signal to the vehicle through a wireless transmission module and provide corresponding solutions.

Images