CN113222187B - Intelligent monitoring method for brake health of shared power-assisted vehicle - Google Patents
Intelligent monitoring method for brake health of shared power-assisted vehicle Download PDFInfo
- Publication number
- CN113222187B CN113222187B CN202110406973.5A CN202110406973A CN113222187B CN 113222187 B CN113222187 B CN 113222187B CN 202110406973 A CN202110406973 A CN 202110406973A CN 113222187 B CN113222187 B CN 113222187B
- Authority
- CN
- China
- Prior art keywords
- data
- vehicle
- brake
- braking
- degree
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 230000036541 health Effects 0.000 title claims abstract description 21
- 238000000034 method Methods 0.000 title claims abstract description 21
- 238000012544 monitoring process Methods 0.000 title claims abstract description 12
- 238000012417 linear regression Methods 0.000 claims description 28
- 238000004364 calculation method Methods 0.000 claims description 16
- 238000004891 communication Methods 0.000 claims description 14
- 230000009471 action Effects 0.000 claims description 13
- 238000007781 pre-processing Methods 0.000 claims description 12
- 238000005070 sampling Methods 0.000 claims description 8
- 238000012216 screening Methods 0.000 claims description 6
- 238000005259 measurement Methods 0.000 claims description 3
- 238000000611 regression analysis Methods 0.000 claims description 2
- 238000012549 training Methods 0.000 claims description 2
- 230000003213 activating effect Effects 0.000 claims 1
- 238000012423 maintenance Methods 0.000 abstract description 9
- 230000005540 biological transmission Effects 0.000 abstract description 5
- 238000010801 machine learning Methods 0.000 abstract description 4
- 230000008520 organization Effects 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 12
- 230000004913 activation Effects 0.000 description 3
- 230000002159 abnormal effect Effects 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 239000013078 crystal Substances 0.000 description 1
- 229910003460 diamond Inorganic materials 0.000 description 1
- 239000010432 diamond Substances 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000010438 heat treatment Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000012886 linear function Methods 0.000 description 1
- 238000010295 mobile communication Methods 0.000 description 1
- 238000004806 packaging method and process Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/20—Administration of product repair or maintenance
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B62—LAND VEHICLES FOR TRAVELLING OTHERWISE THAN ON RAILS
- B62J—CYCLE SADDLES OR SEATS; AUXILIARY DEVICES OR ACCESSORIES SPECIALLY ADAPTED TO CYCLES AND NOT OTHERWISE PROVIDED FOR, e.g. ARTICLE CARRIERS OR CYCLE PROTECTORS
- B62J45/00—Electrical equipment arrangements specially adapted for use as accessories on cycles, not otherwise provided for
- B62J45/40—Sensor arrangements; Mounting thereof
- B62J45/41—Sensor arrangements; Mounting thereof characterised by the type of sensor
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M17/00—Testing of vehicles
- G01M17/007—Wheeled or endless-tracked vehicles
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/40—Business processes related to the transportation industry
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Marketing (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Entrepreneurship & Innovation (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Mechanical Engineering (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Electric Propulsion And Braking For Vehicles (AREA)
- Regulating Braking Force (AREA)
Abstract
The invention provides an intelligent monitoring method for the brake health of a shared power-assisted vehicle, which utilizes the existing hardware facilities of the existing power-assisted vehicle to the greatest extent to invent a method for completing the acquisition and transmission of key data at the vehicle end by installing a brake sensor and utilizing the data reporting and speed detecting module of the existing module. And the platform side completes the association of the vehicle end, the user terminal and other characteristic data and the data classification for machine learning. The machine learning model adopts a supervised learning mode, a monitoring model is built through a large number of shared vehicles and operation data in the city in the initial stage, and a regression model is built through a vehicle data sample with normal braking. When the data is detected, the data to be detected is brought into the model to calculate residual errors, the health state of the brake is detected through residual error threshold judgment, a work order can be sent down to an on-line operation and maintenance work order to be overhauled when the normal performance of the brake cannot be achieved due to the fact that the use loss of the brake is large, and meanwhile, the problem of vehicle organization problem is marked for secondary operation of vehicles.
Description
Technical Field
The invention belongs to the technical field of mobile communication, and particularly relates to an intelligent monitoring method for the brake health of a shared power-assisted vehicle.
Background
The sharing booster vehicle is more and more popular as the last kilometer of urban travel. Along with the continuous heating of riding hot tides, the sharing booster vehicle can certainly remodel a public transportation system, and the industrial ecology of the urban public transportation system is gradually adjusted. However, it should be noted that the operation and maintenance of the shared vehicle is the most important part of the daily operations of the shared vehicle operation platform, and the problem of vehicle failure increasingly affects the experience of the rider. In the big data age, each sharing booster vehicle is a movable organism sensor and a movable data collector, and the problem of difficult operation is solved by adopting a novel technical means.
The shared booster bicycle has the problems of large throwing quantity, difficult operation and maintenance monitoring and the like, and for the current shared booster bicycle operation environment, operators can only monitor the condition of a storage battery according to background data, and the health degree of other main components is reported by means of autonomous discovery of users, and then the operators are arranged for maintenance, so that the operation and maintenance efficiency is low; meanwhile, the driving speed of the shared booster vehicle is high, and under complex road conditions such as urban roads, a driver needs to brake in time when the road conditions change, so that the consequences caused by brake failure can not be estimated. The brake is used as an important guarantee for driver and passenger safety and road traffic safety, and a user can independently find that the reporting and overhauling mode can not adapt to the operation management requirement and the road traffic safety requirement.
Disclosure of Invention
The invention aims to: in order to solve the technical problems in the background art, the invention provides an intelligent monitoring method for the brake health of a shared power-assisted vehicle, which comprises the following steps:
Step1, data acquisition is carried out;
Step 2, preprocessing the collected data;
step 3, performing model training;
and 4, deploying a trained model, and predicting the brake health of the shared moped.
The step 1 comprises the following steps:
Step 1-1, installing an angle sensor at a linear brake expansion bearing of a shared booster vehicle to measure a user braking angle, and adding an information acquisition module in an original module for executing steps 1-2 to 1-4;
step 1-2, obtaining a braking degree; the information acquisition module is activated to carry out information acquisition when the angle sensor is activated, the acquisition is closed when the braking is finished, and strict clock synchronization is needed during data acquisition so as to ensure the reliability and effectiveness of establishing linear regression model data. The degree of braking is expressed in percent and is calculated in two steps as follows:
the first step, the braking angle obtained in step 1-1 is divided by the maximum braking holding angle (the shared vehicle has the characteristics of large throwing quantity and high standardization degree, so that the maximum braking holding angle can be measured in advance), the braking degree in one acquisition period is obtained, and the braking degree A in one acquisition period is calculated by adopting the following formula:
wherein B represents a measurement angle, and C represents a Max angle;
The information acquisition module finally outputs an average value of the braking held by the user in one braking action;
secondly, calculating an average value D of the degree of holding the wire brake by the user in one braking action by adopting the following formula:
Wherein E represents the braking degree of one cycle of the front brake, F represents the braking degree of one cycle of the rear brake, and G represents the braking duration;
a sampling frequency H (the comprehensive data size and the complexity suggest 5 acquisitions per second) is preset, the frequency is used as the basis for calculating the braking time, the braking degree and the deceleration, and the calculation formula of the sampling interval time T is as follows:
T＝1/H
The calculation formula of the brake duration G is as follows:
G＝I*T
Wherein I represents the number of effective acquisitions;
step 1-3, calculating the vehicle deceleration A avg by adopting a common clock under the premise of synchronizing with the brake degree information data:
Wherein: v 1 denotes the initial speed, V 2 denotes the final speed;
step 1-4, when the user brake operation happens, wake up the information acquisition module through setting up a logic AND gate decision circuit, calculate the braking degree through carrying out step 1-2, calculate the vehicle deceleration through carrying out step 1-3, after the user brake action is finished, angle sensor does not give the information acquisition module enable signal, the information acquisition module packs the car end data that obtain of gathering, transmit to current communication module through the data bus, the communication module relies on communication chip and the SIM card in the intelligent lock to the basic station transmission data in the service area, the platform side receives car end data through the operator network. The vehicle end data comprise braking time, braking degree and deceleration data.
The step 2 comprises the following steps: the operation platform carries out the first data association backfill on the data through the journey number and the vehicle ID information, and completes association of the uploaded data and the user terminal data; then determining road section information during vehicle braking operation according to the time offset, and inquiring through a comprehensive database to obtain road surface information, weather information of an operation city at the time and vehicle basic information; after the two times of data association operation, an original data set for learning is obtained, and the original data set finishes data screening and classification (labeled data) according to weather conditions and vehicle braking states.
The step 3 comprises the following steps:
Step 3-1, data screening: firstly removing vehicle data with poor braking conditions according to existing vehicle quality label data and weather label data, then carrying out regression analysis on an original data set according to different weather data labels, and removing data with sample variance more than 0.1 in the original data set, so that the model is more accurate in building;
Step 3-2, adopting a linear regression model, setting an input value of the model as x, wherein x is characteristic data, an output value is a target value Y, and Y is a predicted value;
The input feature x in the model is a multidimensional vector, x 1 represents the braking degree of the ith braking action in the original data set, x 2 represents the vehicle deceleration obtained by the ith braking action in the original data set, and the following formula is obtained:
Y＝θ0+θ1x1+θ2x2
Where θ 0 is the bias term, θ 1 is the brake degree weight, and θ 2 is the vehicle deceleration weight;
Converting the linear regression equation y=θ 0+θ1x1+θ2x2 to a linear regression equation y=xa+b, a representing the slope, B representing the intercept, x representing the eigenvalue, when there is only one independent variable x, a and B are calculated according to the following equation:
Where x and y are sample values in the data set with the normal tag of brake obtained by data preprocessing, AndThe average value of each sample in the original data set with the normal brake tag is obtained through data preprocessing; that is, x=avg (Y1) and y=avg (Y2), Y1 represents a known brake level data set, and Y2 represents a known vehicle deceleration data set.
Step 4 comprises:
Step 4-1, for the shared power-assisted vehicle brake data needing to be predicted, firstly bringing the average brake degree x of the whole vehicle into a linear regression equation, returning a predicted Y value by the linear regression equation, comparing the predicted Y value with the actual deceleration of the vehicle, subtracting the actual deceleration of the vehicle from the predicted Y value, and continuing the step 4-2 if the obtained result is negative, wherein the positive result indicates that the brake is normal, and not judging;
Step 4-2, judging that the braking is normal if the residual error Y 2 is smaller than 0.05, judging that the vehicle braking health is poor when the residual error is more than 0.1 and more than or equal to 0.05, and needing to be overhauled; and stopping the operation of the vehicle if the residual error Y 2 is more than or equal to 0.1.
The beneficial effects are that: the intelligent lock of the moped is integrated with an internet of things communication module (data reporting and instruction receiving), a speed detection module (overspeed reminding) module and a positioning module. The method utilizes the existing hardware facilities of the existing moped to the greatest extent to invent a device for completing the acquisition and transmission of key data at the vehicle end by installing a brake sensor and utilizing the data reporting and speed detecting module of the existing module. And the platform side completes the association of the vehicle end, the user terminal and other characteristic data and generates the characteristic data for machine learning. The machine learning model adopts a supervised learning mode, a monitoring model is built through a large number of shared vehicles and operation data in a city at an initial stage, a regression mode is utilized to fit a brake health degree straight line, the health state of a brake is detected through threshold setting, a work order can be sent down to an operation and maintenance work order to overhaul when the brake is used and damaged greatly and the normal performance of the brake is not achieved, and meanwhile, the problem vehicle is marked for organizing the problem vehicle secondary operation. The scheme utilizes the control system of the existing booster vehicle and the platform of the operation company to provide an active monitoring method for the health degree of the brake, so as to actively find out the failure of the brake, remind operation and maintenance personnel to overhaul, prevent the problem vehicle from operating secondarily and ensure that the shared booster vehicle is safer.
Drawings
The foregoing and/or other advantages of the invention will become more apparent from the following detailed description of the invention when taken in conjunction with the accompanying drawings and detailed description.
FIG. 1 is a schematic diagram of a business process according to the present invention
FIG. 2 is a schematic diagram of the overall architecture of the present invention.
FIG. 3 is a schematic diagram of the present situation of a wire-type brake for a two-wheeled vehicle.
FIG. 4 is a schematic view of the installation position of a linear brake sensor of a two-wheeled vehicle.
Fig. 5 is a schematic diagram of the sensor measurement principle.
FIG. 6 is a flow chart of the layout of the vehicle-side data acquisition module and the data preprocessing.
Fig. 7 is a schematic diagram of a conventional shared vehicle communication scheme.
Fig. 8 is a schematic diagram of a data feature extraction flow.
Fig. 9 is a schematic diagram of the data screening principle.
Fig. 10 is a schematic of a linear regression equation fit.
FIG. 11 is a schematic diagram of a data prediction model.
Detailed Description
The conditions affecting the braking conditions are the friction between the road and the tires (road type, weather, vehicle load, tire wear degree), the braking force of the brakes, and in order to keep road traffic safety, the vehicles need to keep good braking performance under any conditions, so that the detection of the braking force of the brakes is critical.
The most visual manifestation of the brake operating state is the deceleration that the vehicle acquired when the brake was started. Under the condition of brake failure or sensitivity reduction, riding is performed manually to obtain deceleration, and a wire brake is required to be held for a long time to obtain corresponding deceleration; on the contrary, when a larger deceleration is obtained in unit time, namely, the brake works normally, it can be seen that a very strong linear relationship exists among the holding degree of the wire brake, the deceleration obtained by the vehicle and the brake duration. The present invention establishes a regression model based on user behavior by detecting the degree of grip of the vehicle's brake, the deceleration achieved by the vehicle, and the duration of braking to identify the problem vehicle. The business flow chart is shown in figure 1;
The invention realizes the brake health monitoring of the shared moped based on NB-IoT and a supervised learning environment, and relates to vehicle end data acquisition, information transmission, data integration, model establishment and prediction; five parts are formed, and the whole structure diagram is shown in figure 2;
1. data acquisition and front-end preprocessing
According to the model requirement, the vehicle end data acquisition module is required to record three types of information, namely duration time of each brake, brake degree and vehicle deceleration. Because the data has higher formatting degree, the data required by the model is preprocessed at the front end and then returned to the platform, and the data acquisition method and the data processing method are summarized respectively.
(1) Acquisition of time information
The existing crystal oscillator clock source in the intelligent lock installed on the existing shared vehicle can acquire the time synchronization signal and the timing reference signal from the existing clock source or add the time synchronization signal and the timing reference signal newly.
(2) Data acquisition of brake degree (as shown in fig. 3, 4 and 5):
the shared vehicle has the characteristics of large throwing quantity and high standardization degree, so that the maximum holding angle (Max angle can be measured in advance) of the brake is realized.
(3) Acquisition of deceleration information
The deceleration is taken as an important index for evaluating the braking performance, the information such as the speed per hour, the time and the like of the vehicle is required to be acquired by measuring the deceleration of the vehicle, the deceleration is acquired by adopting an average deceleration method, and the deceleration calculation formula is as follows:
wherein: v units are m/s; t is s
The speed information can be finished through the speed measuring module of the existing overspeed alarm module of the vehicle, only the data synchronization (time synchronization) with the braking degree information is required to be finished, and the time synchronization is finished through the adoption of a common clock.
(4) Front-end data preprocessing and data acquisition implementation mode
When the brake is started, the angle sensor acquires a signal, the activation judgment circuit wakes up the information acquisition module, and the information acquisition module starts to work. The front-end information acquisition module acquires four-dimensional information of front and rear braking degrees, braking time, initial speed and final speed, wherein the front and rear braking degrees are acquired by means of angle sensors arranged at the online brake bearing, and the angle information is acquired and then is processed by the module according to a braking degree calculation formula 1 (described above) to obtain the braking degree in one acquisition period; the braking time is recorded from the time of module activation, the initial braking speed is the first data recorded during module activation, the final speed is the data of the previous period acquired during the module enabling signal loss, and the schematic diagram of the front-end information acquisition module is shown in fig. 6.
1) Calculation of brake time
The module needs strict clock synchronization to ensure the reliability and effectiveness of establishing linear regression model data, and needs to preset a sampling frequency (the comprehensive data size and complexity suggest 5 acquisitions per second), and the frequency is used as the basis for calculating the braking time, the braking degree and the deceleration, and the calculation formula is as follows:
Sampling interval time (sec) =1 (sec)/sampling frequency (sec/time)
The calculation formula of the braking time is as follows:
Brake duration (seconds) =effective acquisition times (times) ×sampling interval (seconds)
2) Calculation of the degree of braking output by the module
The module finally outputs an average value of the brake held by the user in one braking action, and the calculation formula is as follows:
3) Deceleration rate
The information of the speed per hour, time and the like of the vehicle is required to be acquired when the deceleration of the vehicle is measured, the deceleration is acquired by adopting an average deceleration method, and the deceleration calculation formula is as follows:
wherein: v1 initial velocity V2 final velocity, unit is m/s; t brake time is s;
The user terminal can report the journey information to the operation platform in real time, so that front-end collection is not needed.
4. Data reporting operation
After each time of braking is finished, the information acquisition module completes data packaging operation, the data packet is transmitted to the communication module through the data bus, the communication module transmits data to the base station in the service area by means of the communication chip and the SIM card in the intelligent lock, and the platform side receives vehicle end data through the operator network.
The data reporting operation is completed by means of the existing intelligent lock internal communication module, and a communication system diagram of the intelligent lock internal communication module is shown in fig. 7;
The brake data is reported according to the travel section, and the uploading flow of the existing vehicle data is adapted.
5. Data association and preprocessing
The vehicle reporting data and the terminal data are reported to an operation platform through a transmission network, the platform carries out first data association backfilling on the data through information such as a journey number, a vehicle ID and the like, and the association of the vehicle end data and the user terminal data is completed; then determining road section information during vehicle braking operation according to the time offset, and inquiring and acquiring key advanced parameters such as road surface information, weather information (determining whether a road surface is slippery) and vehicle basic information of an operation city at the time through a comprehensive database; and obtaining the original data for learning after two times of data association operation. The characteristic engineering flow chart is shown in fig. 8;
6. Model building, learning and prediction
According to the business objective, the final objective of the model is to predict and check the brake health of the operating vehicle through the collection of a plurality of pieces of front-end information, a regression mode can be utilized to fit a linear regression equation of a normal braking vehicle, and after the model is established, the model can judge whether the new instance is normal or abnormal by calculating the variance of the new instance and the linear regression equation when the new instance is seen. The health state of the brake is detected through the setting of the variance threshold, a work order can be sent to a first-line operation and maintenance work order for overhauling when the normal performance of the brake cannot be achieved due to the fact that the use loss of the brake is large, and meanwhile, the problem of vehicle organization problem vehicles is marked for secondary operation.
1) Data screening
When the model is initially built, special data (the braking degree is small, but the working condition is large when the deceleration is obtained during the uphill running braking) with high discrete value and problem vehicle data with poor braking working condition are removed by means of existing vehicle good-bad label data and weather label data (a data set of distinguishing the vehicle braking good-bad on basic data), and the vehicle with normal braking is used for returning modeling according to weather conditions. In the step, a scattered point schematic diagram is constructed by using the average braking degree and the actual deceleration of the whole vehicle, and is shown in figure 9;
2) Removing problem sample and special sample for fitting
The data screened in the previous step is predicted by adopting a linear regression model, the input value of the model is set as x (characteristic data), and the output value is set as a target value Y (predicted value). In the present model, the input feature x is a multidimensional vector, for example, x1 represents the braking degree of the ith braking action in the data set, x2 represents the vehicle deceleration obtained by the ith braking action in the data set, so that the input feature x and the deceleration y can be assumed to satisfy a linear function, and the formula is as follows:
Y＝θ0+θ1x1+θ2x2
Wherein Y is a predicted value, θ 0 is a bias term, θ 1 is a braking degree weight, and θ 2 is a vehicle deceleration weight
The purpose of linear regression is to: the value of theta is calculated, an optimal value of theta is selected to form an algorithm formula, a linear regression formula Y=theta 0+θ1x1+θ2x2 is converted into a linear equation y=xA+B (Y target value (vehicle deceleration), A slope, B intercept and x characteristic value (user braking degree)), the linear regression equation is obtained by adopting a least square method, when the linear regression model is optimal, the difference between the predicted value and the actual value of all samples is minimized, and the difference between the predicted value and the actual value has positive and negative properties, so that the squared value is required to be minimized (the high fitting degree of the model is obtained), and the discrete degree of data determines the accuracy of the regression equation calculated by the least square method. When there is only one argument x, a and B in the linear regression equation are calculated according to the following formula:
In the formula, x and y are sample values in a data set with a normal brake tag obtained through data preprocessing, And/>The average value of each sample in the data set with the normal brake tag is obtained through data preprocessing; that is, x=avg (Y1) and y=avg (Y2), Y1 represents a known brake level data set, Y2 represents a known vehicle deceleration data set, AVG is an averaging function.
The model is built by adopting a normal brake sample, and a calculation schematic image is shown in fig. 10;
In fig. 10, the horizontal axis (X) represents the degree of braking of the vehicle, and the vertical axis (Y) represents the actual deceleration of the vehicle. In the figure, diamond points represent real Y values, square points represent best fit Y values obtained by least square calculation, and Y values are connected to form a linear regression equation y=xA+B obtained by the model.
3) Model prediction
After the regression model is built by adopting normal data, when the model sees a new instance, whether the new instance is normal or abnormal can be judged by calculating the residual errors of the new instance and the regression equation on the plane rectangular coordinate system. The broken line in the lower graph is a regression equation established by adopting a normal sample, the data with low deceleration and high braking degree can be seen as problem data, and a calculation schematic image is shown in fig. 11;
the specific research and judgment process is as follows:
1. A linear regression equation y=xa+b (predictive model) derived by the above two steps;
2. When the new data set is researched and judged, the average braking degree (x) of the whole vehicle is firstly brought into a linear regression equation, and the linear regression equation returns a predicted Y value, and the formula is as follows:
y＝xA+B
y is a predictive function value, A is a slope (known by a linear regression equation model), x is an average braking degree (independent variable) of the whole vehicle, and B is an intercept (known by the linear regression equation model).
3. Comparing the predicted Y value with the actual deceleration of the vehicle, wherein the predicted Y value-the actual deceleration value is negative, namely, the value is under the image of the linear regression equation, the value is normal in regular braking, and the judgment is not carried out, and the judgment rule is as follows:
IF (true vehicle deceleration-predicted Y value <0, lower equation, upper equation)
4. And judging the residual error size of the sample with the value which does not meet the model requirement, and identifying the brake health degree. Sample variance is adopted in the model for studying and judging, namely, the Y value obtained by the formula is squared, and the formula is as follows:
Residual = Y 2
If the residual error Y 2 is smaller than 0.05, the brake is considered normal, and if the residual error is larger than 0.1 and larger than or equal to Y 2 and larger than or equal to 0.05, the vehicle brake is considered to be bad in health degree, and maintenance is needed; and stopping the operation of the vehicle when the residual error Y 2 is more than or equal to 0.1.
The invention provides an intelligent monitoring method for the brake health of a shared power-assisted vehicle, which has a plurality of methods and approaches for realizing the technical scheme, the above description is only a preferred embodiment of the invention, and it should be pointed out that a plurality of improvements and modifications can be made by one of ordinary skill in the art without departing from the principle of the invention, and the improvements and modifications are also considered as the protection scope of the invention. The components not explicitly described in this embodiment can be implemented by using the prior art.
Claims (4)
1. The intelligent monitoring method for the brake health of the shared power-assisted vehicle is characterized by comprising the following steps of:
Step1, data acquisition is carried out;
Step 2, preprocessing the collected data;
step 3, performing model training;
Step 4, deploying a trained model, and predicting the brake health of the shared moped;
The step 1 comprises the following steps:
Step 1-1, installing an angle sensor at a linear brake expansion bearing of a shared booster vehicle to measure a user braking angle, and adding an information acquisition module in an original module for executing steps 1-2 to 1-4;
Step 1-2, obtaining a braking degree; activating an information acquisition module to acquire information when the angle sensor is activated, and closing the acquisition when the braking is finished, wherein the information acquisition module is used for carrying out the calculation in the following two steps:
the first step, the braking degree A in one acquisition period is calculated by adopting the following formula:
wherein B represents a measurement angle, and C represents a Max angle;
The information acquisition module finally outputs an average value of the braking held by the user in one braking action;
secondly, calculating an average value D of the degree of holding the wire brake by the user in one braking action by adopting the following formula:
Wherein E represents the braking degree of one cycle of the front brake, F represents the braking degree of one cycle of the rear brake, and G represents the braking duration;
A sampling frequency H is preset, and a calculation formula of sampling interval time T is as follows:
T＝1/H
The calculation formula of the brake duration G is as follows:
G＝I*T
Wherein I represents the number of effective acquisitions;
step 1-3, calculating the vehicle deceleration A avg by adopting a common clock under the premise of synchronizing with the brake degree information data:
Wherein: v 1 denotes the initial speed, V 2 denotes the final speed;
Step 1-4, when the user brake operation happens, wake up the information acquisition module through setting up a logic AND gate decision circuit, calculate the braking degree through carrying out step 1-2, calculate the vehicle deceleration through carrying out step 1-3, the angle sensor does not give the information acquisition module enable signal after the user brake action is finished, the information acquisition module packs the acquired vehicle end data, transmits the data to the existing communication module, the communication module transmits the data to the base station in the service area by means of the communication chip and the SIM card in the intelligent lock, and the platform side receives the vehicle end data through the operator network; the vehicle end data comprise braking time, braking degree and deceleration data.
2. The method of claim 1, wherein step 2 comprises: the operation platform carries out the first data association backfill on the data through the journey number and the vehicle ID information, and associates the uploaded data with the user terminal data; then determining road section information during vehicle braking operation according to the time offset, and inquiring through a comprehensive database to obtain road surface information, weather information of an operation city at the time and vehicle basic information; after the two times of data association operation, an original data set for learning is obtained, and the original data set is subjected to data screening and classification according to weather conditions and vehicle braking states.
3. The method of claim 2, wherein step 3 comprises:
step 3-1, data screening: removing vehicle data with poor braking conditions by means of existing vehicle quality label data and weather label data, carrying out regression analysis on an original data set according to different weather data labels, and removing data with sample variance larger than 0.1 in the original data set;
Step 3-2, adopting a linear regression model, setting an input value of the model as x, wherein x is characteristic data, an output value is a target value Y, and Y is a predicted value;
The input feature x in the model is a multidimensional vector, x 1 represents the braking degree of the ith braking action in the original data set, x 2 represents the vehicle deceleration obtained by the ith braking action in the original data set, and the following formula is obtained:
Y＝θ0+θ1x1+θ2x2
Where θ 0 is the bias term, θ 1 is the brake degree weight, and θ 2 is the vehicle deceleration weight;
Converting the linear regression equation y=θ 0+θ1x1+θ2x2 to a linear regression equation y=xa+b, a representing the slope, B representing the intercept, x representing the eigenvalue, when there is only one independent variable x, a and B are calculated according to the following equation:
Where x and y are sample values in the data set with the normal tag of brake obtained by data preprocessing, And/>The average value of each sample in the original data set with the normal brake tag obtained through data preprocessing, namely x=avg (Y1) and y=avg (Y2), wherein Y1 represents a known brake degree data set, and Y2 represents a known vehicle deceleration data set.
4. A method according to claim 3, wherein step 4 comprises:
Step 4-1, for the shared power-assisted vehicle brake data needing to be predicted, firstly bringing the average brake degree x of the whole vehicle into a linear regression equation, returning a predicted Y value by the linear regression equation, comparing the predicted Y value with the actual deceleration of the vehicle, subtracting the predicted Y value from the actual deceleration of the vehicle, and continuing the step 4-2 if the obtained result is negative, wherein the positive result indicates that the brake is normal, and not judging;
step 4-2, judging that the braking is normal if the residual error Y2 is less than 0.05, judging that the vehicle braking health is poor when the residual error is more than or equal to 0.1 and more than or equal to 0.05, and needing to be overhauled; and stopping the operation of the vehicle if the residual error Y2 is more than or equal to 0.1.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110406973.5A CN113222187B (en) | 2021-04-15 | 2021-04-15 | Intelligent monitoring method for brake health of shared power-assisted vehicle |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110406973.5A CN113222187B (en) | 2021-04-15 | 2021-04-15 | Intelligent monitoring method for brake health of shared power-assisted vehicle |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113222187A CN113222187A (en) | 2021-08-06 |
| CN113222187B true CN113222187B (en) | 2024-05-03 |
Family
ID=77087419
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110406973.5A Active CN113222187B (en) | 2021-04-15 | 2021-04-15 | Intelligent monitoring method for brake health of shared power-assisted vehicle |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113222187B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114254766B (en) * | 2021-12-10 | 2024-07-05 | 智己汽车科技有限公司 | Method and system for jointly training machine learning model at two ends of vehicle cloud |
| CN117556874B (en) * | 2023-12-26 | 2024-03-12 | 宝鸡中铁中联科技发展有限公司 | Vehicle control method and system based on multi-source data |
| CN117804794B (en) * | 2023-12-29 | 2024-10-18 | 无锡信贝科技发展有限公司 | Brake performance detection and early warning method and system for power-assisted bicycle |
| CN117649213B (en) * | 2024-01-30 | 2024-04-19 | 四川宽窄智慧物流有限责任公司 | Front-end management method and system for transportation safety |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106056140A (en) * | 2016-05-27 | 2016-10-26 | 大连楼兰科技股份有限公司 | System and method for establishing working condition division remote damage assessment of different vehicle types based on artificial intelligence supervised learning linear regression method |
| CN108182798A (en) * | 2017-12-29 | 2018-06-19 | 宋孟袁 | Public bicycles vehicle condition monitors system |
| CN209356189U (en) * | 2019-03-28 | 2019-09-06 | 吉林大学 | A kind of brake data acquisition device, the vehicles and tire for vehicles |
| CN111177939A (en) * | 2020-01-03 | 2020-05-19 | 中国铁路郑州局集团有限公司科学技术研究所 | Deep learning-based brake cylinder pressure prediction method for train air brake system |
| CN111497812A (en) * | 2019-01-31 | 2020-08-07 | 天津所托瑞安汽车科技有限公司 | Vehicle braking system monitoring method based on platform |
| WO2021019978A1 (en) * | 2019-07-31 | 2021-02-04 | 日信工業株式会社 | Operation amount detection device for handlebar vehicle |
| CN112550262A (en) * | 2020-12-11 | 2021-03-26 | 奇瑞汽车股份有限公司 | Health state diagnosis method and system for automobile brake system |
-
2021
- 2021-04-15 CN CN202110406973.5A patent/CN113222187B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106056140A (en) * | 2016-05-27 | 2016-10-26 | 大连楼兰科技股份有限公司 | System and method for establishing working condition division remote damage assessment of different vehicle types based on artificial intelligence supervised learning linear regression method |
| CN108182798A (en) * | 2017-12-29 | 2018-06-19 | 宋孟袁 | Public bicycles vehicle condition monitors system |
| CN111497812A (en) * | 2019-01-31 | 2020-08-07 | 天津所托瑞安汽车科技有限公司 | Vehicle braking system monitoring method based on platform |
| CN209356189U (en) * | 2019-03-28 | 2019-09-06 | 吉林大学 | A kind of brake data acquisition device, the vehicles and tire for vehicles |
| WO2021019978A1 (en) * | 2019-07-31 | 2021-02-04 | 日信工業株式会社 | Operation amount detection device for handlebar vehicle |
| CN111177939A (en) * | 2020-01-03 | 2020-05-19 | 中国铁路郑州局集团有限公司科学技术研究所 | Deep learning-based brake cylinder pressure prediction method for train air brake system |
| CN112550262A (en) * | 2020-12-11 | 2021-03-26 | 奇瑞汽车股份有限公司 | Health state diagnosis method and system for automobile brake system |
Non-Patent Citations (3)
| Title |
|---|
| An Anti-Lock Braking System for Bicycles;Matteo Corno等;2018 IEEE Conference on Control Technology and Applications (CCTA);834-839 * |
| 电动车部件损耗监测及反馈系统的实现方法研究;张磊等;计算机产品与流通;20181215(12);146-147 * |
| 自行车制动力测试及其设备;胡若平等;检验检疫科学(第Z1期);20-22 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113222187A (en) | 2021-08-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113222187B (en) | Intelligent monitoring method for brake health of shared power-assisted vehicle | |
| CN111257415B (en) | Tunnel damage detection management system based on mobile train vibration signal | |
| CN102881162B (en) | Data processing and fusion method for large-scale traffic information | |
| CN102622798B (en) | Passenger flow statistical analysis system | |
| CN100538761C (en) | Built-in intelligent fault diagnosing device and method based on the data fusion pattern-recognition | |
| WO2016074600A1 (en) | Vehicle operation monitoring, overseeing, data processing and overload monitoring method and system | |
| CN111477005B (en) | Intelligent perception early warning method and system based on vehicle state and driving environment | |
| CN102717765A (en) | Fatigue driving detection method and anti-fatigue driving auxiliary device | |
| CN111260822B (en) | Rail transit vehicle health state analysis method and terminal based on big data | |
| CN102360525A (en) | Discriminant analysis-based high road real-time traffic accident risk forecasting method | |
| CN113752843B (en) | Power battery thermal runaway early warning device and method based on Saybolt physical system | |
| CN102622885B (en) | Method and device for detecting traffic incidents | |
| CN104386449A (en) | Intelligent protection device for online detection of head/tail wheels of mining belt conveyor | |
| CN103692969A (en) | Monitoring method for long time driving | |
| CN112883645A (en) | Single battery voltage inconsistency fault prediction method and device and server | |
| CN108961473A (en) | A kind of vehicle-state assessment method for early warning based on intelligent network connection automobile control centre | |
| CN105539450A (en) | Automatic identification method and device of driving stroke | |
| CN114526930B (en) | Intelligent network-connected automobile fault detection method and system | |
| CN117392855B (en) | Vehicle overrun overload probability identification method and system based on satellite positioning data | |
| CN112129290A (en) | System and method for monitoring riding equipment | |
| CN113156913A (en) | ABS fault diagnosis system and method | |
| CN204359419U (en) | A kind of vehicle strain weighing system | |
| CN104501929A (en) | Vehicle strain weighing system | |
| CN103278603B (en) | Utilize number plate recognition technology monitoring motor vehicle to the system and method for environmental impact | |
| CN212009835U (en) | Traffic big data analysis early warning platform |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |