CN108053075B - Scrapped vehicle prediction method and system - Google Patents
Scrapped vehicle prediction method and system Download PDFInfo
- Publication number
- CN108053075B CN108053075B CN201711445897.9A CN201711445897A CN108053075B CN 108053075 B CN108053075 B CN 108053075B CN 201711445897 A CN201711445897 A CN 201711445897A CN 108053075 B CN108053075 B CN 108053075B
- Authority
- CN
- China
- Prior art keywords
- vehicle
- preset time
- time period
- preset
- module
- 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
- 238000000034 method Methods 0.000 title claims abstract description 25
- 238000012549 training Methods 0.000 claims abstract description 87
- 238000012360 testing method Methods 0.000 claims abstract description 49
- 238000012423 maintenance Methods 0.000 claims description 45
- 238000007781 pre-processing Methods 0.000 claims description 15
- 238000004422 calculation algorithm Methods 0.000 claims description 11
- 238000007637 random forest analysis Methods 0.000 claims description 11
- 238000011156 evaluation Methods 0.000 claims description 9
- 238000000605 extraction Methods 0.000 claims description 7
- 238000010801 machine learning Methods 0.000 abstract description 5
- 238000001914 filtration Methods 0.000 description 3
- 230000003203 everyday effect Effects 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
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
- G06Q10/04—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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- 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
- G06Q50/10—Services
- G06Q50/26—Government or public services
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Human Resources & Organizations (AREA)
- General Physics & Mathematics (AREA)
- Economics (AREA)
- Tourism & Hospitality (AREA)
- Strategic Management (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Marketing (AREA)
- Development Economics (AREA)
- General Business, Economics & Management (AREA)
- Game Theory and Decision Science (AREA)
- Artificial Intelligence (AREA)
- Primary Health Care (AREA)
- Health & Medical Sciences (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- General Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Educational Administration (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Train Traffic Observation, Control, And Security (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a method and a system for predicting a scrapped vehicle, and belongs to the field of intelligent traffic big data. The method comprises the following steps: extracting characteristic information of each vehicle from the track data of each vehicle in a first preset time period; generating training samples according to the characteristic information of each vehicle and a preset scrapped vehicle standard, and dividing the training samples into a training set and a testing set; training a prediction model according to the training set, and testing the prediction model by adopting a test set; and predicting the future scrapped vehicle by using the tested prediction model. According to the method, the training of the prediction model is carried out on the basis of historical track data of each vehicle in a certain period, so that the finally obtained prediction model and the prediction result are more scientific and reliable; meanwhile, the operation rule of the vehicle before scrapping is found out by combining with machine learning, and the problem of hysteresis in the prior art is solved; and the marketing strategy can be determined more accurately for manufacturers of various vehicles.
Description
Technical Field
The invention relates to the field of intelligent traffic big data, in particular to a method and a system for predicting a scrapped vehicle.
Background
With the rapid development of social economy, both private cars and logistics transport vehicles are rapidly increasing at an exponential speed; various vehicle manufacturers are also constantly producing and selling vehicles of various brands, various functions and various sizes. For each vehicle manufacturer, timely understanding of the requirements of vehicle purchasers can enable the vehicle manufacturers to produce more popular vehicles; meanwhile, the scrapped vehicles on the market can be predicted in time, so that the clear marketing requirement is equally important for vehicle manufacturers. The existing prediction method of the scrapped vehicle only considers the number of days for which the vehicle is not operated and sets scrapping rules by using subjective experience; however, whether the vehicle is scrapped or not is judged by the number of days the vehicle is not running, and for vehicle manufacturers, the best marketing opportunity is lost; meanwhile, the rejection rule is set by using subjective experience, so that the prediction result is too subjective and has no convincing power, and a certain hysteresis problem exists. Therefore, how to accurately and scientifically predict the scrapped vehicles still is an urgent problem to be solved for manufacturers of various vehicles.
Disclosure of Invention
In order to overcome the defects in the prior art, the invention provides a method and a system for predicting a scrapped vehicle.
In one aspect, the present invention provides a method for predicting a scrapped vehicle, including:
step S1: extracting characteristic information of each vehicle from the track data of each vehicle in a first preset time period;
step S2: generating training samples according to the characteristic information of each vehicle and a preset scrapped vehicle standard, and dividing the training samples into a training set and a test set;
step S3: training a prediction model according to the training set, and testing the prediction model by adopting the test set;
step S4: and predicting the future scrapped vehicle by using the tested prediction model.
Optionally, before the step S1, the method further includes: preprocessing the track data of each vehicle in a first preset time period;
correspondingly, the step S1 specifically includes: and extracting characteristic information of each vehicle from the preprocessed track data.
Optionally, the trajectory data comprises: a series of position data and driving speed;
the step S1 specifically includes:
dividing the first preset time period into M second preset time periods, and taking the first n second preset time periods as third preset time periods according to the time sequence, wherein n is less than M;
according to the position data and the running speed contained in the track data of each vehicle, counting the number of operation days, the number of parking times of the maintenance station and the parking duration of each vehicle in each second preset time period;
and determining the change trend of the number of the operating days, the number of times of parking at the maintenance station and the average parking duration of each vehicle in the third preset time period according to the number of the operating days, the number of times of parking at the maintenance station and the parking duration of each vehicle in each second preset time period.
And determining the shutdown condition of each vehicle in each second preset time period according to the operation days of each vehicle in each second preset time period and the preset shutdown standard.
Optionally, the step S2 specifically includes:
step S2-1: determining scrapped vehicles and running vehicles in the vehicles according to the shutdown conditions of the vehicles in second preset time periods and preset scrapped vehicle standards;
step S2-2: establishing a corresponding relation between the operation day number change trend, the number of times of parking at a maintenance station and the average parking duration of each vehicle in the scrapped vehicles in the third preset time period and the corresponding vehicle, and using the corresponding relation as a positive sample to obtain a positive sample set; establishing a corresponding relation between the operation day number change trend, the number of times of parking at the maintenance station and the average parking duration of each vehicle in the running vehicles in the third preset time period and the corresponding vehicle, and using the corresponding relation as a negative sample to obtain a negative sample set;
step S2-3: selecting a positive sample from the positive sample set according to a first preset proportion, and selecting a negative sample from the negative sample set to obtain a preset number of training samples;
step S2-4: and dividing the training samples of the preset number into a training set and a test set according to a second preset proportion.
Optionally, the step S3 specifically includes:
step S3-1: training the training set according to a random forest algorithm to obtain a prediction model;
step S3-2: evaluating the prediction model according to the working characteristic curve of the subject and adjusting the model parameters, and outputting the corresponding prediction model when the first area under the working characteristic curve of the subject meets a preset threshold value;
step S3-3: calculating a second area of the output prediction model under a corresponding subject work characteristic curve on the test set, and executing step S4 when the second area meets the preset threshold; otherwise, returning to the step S3-1;
correspondingly, the step S4 specifically includes: and counting the change trend of the operation days, the parking times of the maintenance station and the average parking duration of each vehicle in the third preset time period forwards by taking the current time as the cutoff time, and predicting the future scrapped vehicles and the scrapping probability by using the tested prediction model.
In another aspect, the present invention provides a system for predicting a scrapped vehicle, including:
the extraction module is used for extracting the characteristic information of each vehicle from the track data of each vehicle in a first preset time period;
the generating module is used for generating training samples according to the characteristic information of each vehicle extracted by the extracting module and a preset scrapped vehicle standard;
the dividing module is used for dividing the training samples generated by the generating module into a training set and a test set;
the training module is used for training a prediction model according to the training set divided by the dividing module;
the testing module is used for testing the prediction model trained by the training module by adopting the testing set divided by the dividing module;
and the prediction module is used for predicting the future scrapped vehicle by using the prediction model tested by the test module.
Optionally, the system further comprises: a preprocessing module;
the preprocessing module is used for preprocessing the track data of each vehicle in a first preset time period;
the extraction module is specifically configured to: and extracting the characteristic information of each vehicle from the track data preprocessed by the preprocessing module.
Optionally, the trajectory data comprises: a series of position data and driving speed;
the extraction module is specifically configured to:
dividing the first preset time period into M second preset time periods, and taking the first n second preset time periods as third preset time periods according to the time sequence, wherein n is less than M;
according to the position data and the running speed contained in the track data of each vehicle, counting the number of operation days, the number of parking times of the maintenance station and the parking duration of each vehicle in each second preset time period;
and determining the change trend of the number of the operating days, the number of times of parking at the maintenance station and the average parking duration of each vehicle in the third preset time period according to the number of the operating days, the number of times of parking at the maintenance station and the parking duration of each vehicle in each second preset time period.
And determining the shutdown condition of each vehicle in each second preset time period according to the operation days of each vehicle in each second preset time period and the preset shutdown standard.
Optionally, the generating module specifically includes: determining a submodule, a sample generation submodule and a selection submodule;
the determining submodule is used for determining the scrapped vehicles and the running vehicles in the vehicles according to the shutdown conditions of the vehicles in the second preset time period and the preset scrapped vehicle standards extracted by the extracting module;
the sample generation submodule is used for establishing a corresponding relation between the operation day number change trend, the number of times of parking at the maintenance station and the average parking duration of each vehicle in the scrapped vehicles determined by the determination submodule in the third preset time period and the corresponding vehicle, and using the corresponding relation as a positive sample to obtain a positive sample set; establishing a corresponding relation between the operation day number change trend, the number of times of parking at the maintenance station and the average parking duration of each vehicle in the running vehicles in the third preset time period, which are determined by the determination sub-module, and the corresponding vehicle, and taking the corresponding relation as a negative sample to obtain a negative sample set;
the selecting submodule is used for selecting a positive sample from the positive sample set obtained by the sample generating submodule according to a first preset proportion, and selecting a negative sample from the negative sample set obtained by the sample generating submodule to obtain a preset number of training samples;
the dividing module is specifically configured to: and dividing the training samples of the preset number obtained by the selection submodule into a training set and a test set according to a second preset proportion.
Optionally, the training module specifically includes: a training submodule and an evaluation submodule;
the training submodule is used for training the training set divided by the dividing module according to a random forest algorithm to obtain a prediction model;
the evaluation submodule is used for evaluating the prediction model obtained by the training submodule according to the working characteristic curve of the subject and adjusting the model parameters, and when the first area under the working characteristic curve of the subject meets a preset threshold value, the corresponding prediction model is output;
the test module is specifically configured to: calculating a second area of the prediction model output by the evaluation sub-module under a corresponding test subject working characteristic curve on the test set divided by the dividing module, and triggering the prediction module when the second area meets the preset threshold;
the prediction module is specifically configured to: and counting the change trend of the operation days, the parking times of the maintenance station and the average parking duration of each vehicle in the third preset time period forwards by taking the current time as a cut-off time, and predicting the future scrapped vehicle and the scrapping probability by using the prediction model tested by the test module.
The invention has the advantages that:
according to the method, the generation of the prediction model is carried out on the basis of historical track data of each vehicle in a certain period, the length and the width of the data are greatly increased, and the finally obtained prediction model and the prediction result are more scientific and reliable; meanwhile, the method is combined with machine learning, the operation rule of the vehicle before scrapping is found out according to a random forest algorithm, so that a prediction model for predicting scrapped vehicles and scrapping probability is obtained, and the problem of hysteresis in the prior art is solved; and the marketing strategy can be determined more accurately for manufacturers of various vehicles.
Drawings
Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The drawings are only for purposes of illustrating the preferred embodiments and are not to be construed as limiting the invention. Also, like reference numerals are used to refer to like parts throughout the drawings. In the drawings:
FIG. 1 is a flow chart of a method for predicting a scrapped vehicle according to the present invention;
FIG. 2 is a block diagram of a scrap vehicle forecasting system according to the present invention.
Detailed Description
Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.
Example one
According to an embodiment of the present invention, there is provided a method for predicting a scrapped vehicle, as shown in fig. 1, including:
step 101: extracting characteristic information of each vehicle from the track data of each vehicle in a first preset time period;
preferably, in the present invention, each vehicle is equipped with a vehicle-mounted device, and the vehicle-mounted device reports the trajectory data of the corresponding vehicle at preset time intervals (for example, 30 seconds); trajectory data includes, but is not limited to: a series of position data and driving speed;
in this embodiment, before step 101, the method further includes: preprocessing the track data of each vehicle in a first preset time period;
specifically, the following operations are performed on the trajectory data of each vehicle in a first preset time period:
filtering error data in the track data;
correcting the shifted position data in the trajectory data;
correcting supplementary transmission data in the track data according to the time dimension;
wherein, error data in the filtering track data, such as position information error and driving speed error; correcting the offset position data in the track data, specifically correcting the offset position data by methods such as Fourier filtering and the like;
correspondingly, step 101 specifically includes: and extracting characteristic information of each vehicle from the preprocessed track data.
Further, in this embodiment, step 101 specifically includes:
dividing the first preset time period into M second preset time periods, and taking the first n second preset time periods as third preset time periods according to the time sequence, wherein n is less than M;
according to the position data and the running speed contained in the track data of each vehicle, counting the number of operation days, the number of parking times of the maintenance station and the parking duration of each vehicle in each second preset time period;
and determining the change trend of the number of the operating days, the number of times of parking at the maintenance station and the average parking duration of each vehicle in a third preset time period according to the number of the operating days, the number of times of parking at the maintenance station and the parking duration of each vehicle in each second preset time period.
And determining the shutdown condition of each vehicle in each second preset time period according to the operation days of each vehicle in each second preset time period and the preset shutdown standard.
For example, in this embodiment, the first preset time period is 2016 year 1 month to 2017 year 6 months, the duration of the second preset time period is one month, and the third preset time period is the first 12 months, so that the first preset time period is divided into 12 months, and the third preset time period is 2016 year 1 month to 2016 year 12 months.
Wherein, according to the position data and the running speed contained in the track data of each vehicle, the number of operation days, the number of times of parking at the maintenance station and the parking duration of each vehicle in each second preset time period are counted, specifically:
determining the distance between any two adjacent position data according to the position data, the driving speed and the corresponding time contained in the track data of each vehicle, calculating the driving mileage of each vehicle every day according to the determined distance, taking the corresponding day as the operation time of the vehicle when the driving mileage of each vehicle every day is greater than the preset driving mileage, and counting the operation days of each vehicle in each second preset time period; wherein the preset driving mileage is preferably 5 kilometers;
matching position data contained in the track data of each vehicle with position data of each maintenance station, taking the position data which is successfully matched with the position data of the maintenance station corresponding to the position data as a real parking time, and counting the parking times and parking time of each vehicle in each second preset time period; wherein the preset time is preferably 2 hours.
The method comprises the following steps of determining the change trend of the number of operation days, the number of times of parking at the maintenance station and the average parking duration of each vehicle in a third preset time period according to the number of operation days, the number of times of parking at the maintenance station and the parking duration of each vehicle in each second preset time period, and specifically comprises the following steps: calculating the operation days of each vehicle in each second preset time period contained in the third preset time period by using a least square method to obtain the operation day change trend of each vehicle in the third preset time period; adding the parking times of all vehicle maintenance stations in all second preset time periods contained in the third preset time period to obtain the parking times of all vehicles at the maintenance stations in the third preset time period; averaging the stopping time lengths of all the vehicle maintenance stations in all the second preset time periods contained in the third preset time period to obtain the average stopping time length of all the vehicles at the maintenance stations in the third preset time period;
the preset shutdown standard is preferably, in this embodiment, specifically: when the number of operation days of the vehicle in a second preset time period is zero, recording that the vehicle stops operating in the second preset time period; correspondingly, determining the shutdown condition of each vehicle in each second preset time period according to the number of operation days of each vehicle in each second preset time period and the preset shutdown standard, specifically: and judging whether the operation days of each vehicle in each second preset time period are zero or not, and recording the second preset time period corresponding to the operation days of zero as the corresponding vehicle outage.
Step 102: generating training samples according to the characteristic information of each vehicle and a preset scrapped vehicle standard, and dividing the generated training samples into a training set and a test set;
preferably, in this embodiment, the preset crippled vehicle standard is: the last second preset time period contained in the third preset time period is not stopped, and the operation is stopped in the second preset time periods which are next to the third preset time period, continuous and preset in number; wherein the preset number is preferably 6;
for example, in the present embodiment, a vehicle that was not stopped for 2016 and was stopped for 6 consecutive months (i.e., 1-6 months in 2017) after 2016 12 months is used as a scrapped vehicle.
In this embodiment, step 102 specifically includes:
step 102-1: determining scrapped vehicles and running vehicles in the vehicles according to the shutdown conditions of the vehicles in the second preset time periods and preset scrapped vehicle standards;
step 102-2: establishing a corresponding relation between the operation day number change trend, the number of times of parking at a maintenance station and the average parking duration of each vehicle in the scrapped vehicles in a third preset time period and the corresponding vehicle, and using the corresponding relation as a positive sample to obtain a positive sample set; establishing a corresponding relation between the operation day number change trend, the number of times of parking at a maintenance station and the average parking duration of each vehicle in the running vehicles in a third preset time period and the corresponding vehicle, and taking the corresponding relation as a negative sample to obtain a negative sample set;
step 102-3: selecting a positive sample from the positive sample set according to a first preset proportion, and selecting a negative sample from the negative sample set to obtain a preset number of training samples;
preferably, in this embodiment, the first preset ratio is 1: 1, the preset number is 10000.
Step 102-4: and dividing a preset number of training samples into a training set and a test set according to a second preset proportion.
Preferably, in this embodiment, the second preset ratio is 7: 3.
the standard of the scrapped vehicle is preset according to the requirement and the actual situation.
Step 103: training a prediction model according to the training set, and testing the prediction model by adopting the test set;
in this embodiment, step 103 specifically includes:
step 103-1: training the training set according to a random forest algorithm to obtain a prediction model;
specifically, a Python tool is used for calling a random forest algorithm in the scimit-leann machine learning library to train a training set to obtain a prediction model.
Step 103-2: evaluating the prediction model according to a Receiver operating characteristic Curve (ROC Curve for short) and adjusting model parameters, and outputting a corresponding prediction model when a first area under the Receiver operating characteristic Curve meets a preset threshold;
wherein, the area under the working characteristic curve of the subject, also called AUC, specifically, when the first AUC meets a preset threshold, outputting the corresponding prediction model;
the preset threshold value is preferably 0.8.
Step 103-3: calculating a second area of the output prediction model under a corresponding test subject working characteristic curve on the test set, and executing the step 104 when the second area meets the preset threshold; otherwise, returning to the step 103-1;
in the invention, a random forest algorithm is used for training a prediction model, so that the prediction model is more interpretable and solves the problem of hysteresis in the prior art; meanwhile, the prediction model is tested through the test set, and the accuracy of the prediction model is further guaranteed.
Step 104: and predicting the future scrapped vehicle by using the tested prediction model.
Specifically, the current time is used as the cutoff time, the operation day number change trend, the number of times of parking at the maintenance station and the average parking duration of each vehicle in the third preset time period are counted forwards according to the method, and the tested prediction model is used for predicting the future scrapped vehicle and the scrapping probability.
Example two
According to an embodiment of the present invention, there is provided a system for predicting a scrapped vehicle, as shown in fig. 2, including:
the extraction module 201 is configured to extract feature information of each vehicle from trajectory data of each vehicle within a first preset time period;
the generating module 202 is configured to generate a training sample according to the feature information of each vehicle extracted by the extracting module 201 and a preset scrapped vehicle standard;
a dividing module 203, configured to divide the training samples generated by the generating module 202 into a training set and a test set;
a training module 204, configured to train a prediction model according to the training set divided by the dividing module 203;
the testing module 205 is configured to test the prediction model trained by the training module by using the test set partitioned by the partitioning module 203;
and the prediction module 206 is used for predicting future scrapped vehicles by using the prediction model tested by the testing module 205.
According to an embodiment of the invention, the system further comprises: a preprocessing module;
the preprocessing module is used for preprocessing the track data of each vehicle in a first preset time period;
correspondingly, the extracting module 201 is specifically configured to: and extracting the characteristic information of each vehicle from the track data preprocessed by the preprocessing module.
According to an embodiment of the present invention, the trajectory data includes: a series of position data and driving speed;
correspondingly, the extracting module 201 is specifically configured to:
dividing the first preset time period into M second preset time periods, and taking the first n second preset time periods as third preset time periods according to the time sequence, wherein n is less than M;
according to the position data and the running speed contained in the track data of each vehicle, counting the number of operation days, the number of parking times of the maintenance station and the parking duration of each vehicle in each second preset time period;
and determining the change trend of the number of the operating days, the number of times of parking at the maintenance station and the average parking duration of each vehicle in the third preset time period according to the number of the operating days, the number of times of parking at the maintenance station and the parking duration of each vehicle in each second preset time period.
And determining the shutdown condition of each vehicle in each second preset time period according to the operation days of each vehicle in each second preset time period and the preset shutdown standard.
In this embodiment, the first preset time period, the second preset time period, and the third preset time period may be set according to the requirement, for example, in this embodiment, the first preset time period is 2016 month 1 to 2017 month 6, the second preset time period is one month, and the third preset time period is the first 12 months, the first preset time period is divided into 12 months, and the third preset time period is 2016 month 1 to 2016 month 12.
According to the embodiment of the present invention, the generating module 202 specifically includes: determining a submodule, a sample generating submodule and a selecting submodule, wherein:
the determining submodule is used for determining the scrapped vehicles and the running vehicles in each vehicle according to the shutdown conditions of each vehicle in each second preset time period and the preset scrapped vehicle standard extracted by the extracting module 201;
the preset scrapped vehicle standard is preferably as follows: the last second preset time period contained in the third preset time period is not stopped, and the operation is stopped in the second preset time periods which are next to the third preset time period, continuous and preset in number; wherein the preset number is preferably 6;
for example, in the present embodiment, a vehicle that was not stopped for 2016 and was stopped for 6 consecutive months (i.e., 1-6 months in 2017) after 2016 12 months is used as a scrapped vehicle.
The sample generation submodule is used for establishing a corresponding relation between the operation day number change trend, the number of times of parking at the maintenance station and the average parking duration of each vehicle in the scrapped vehicles in the third preset time period, which are determined by the determination submodule, and the corresponding vehicle, and using the corresponding relation as a positive sample to obtain a positive sample set; establishing a corresponding relation between the operation day number change trend, the number of parking times of the maintenance station and the average parking duration of each vehicle in the running vehicles in the third preset time period, which are determined by the determination sub-module, and the corresponding vehicle, and taking the corresponding relation as a negative sample to obtain a negative sample set;
the selecting submodule is used for selecting a positive sample from a positive sample set obtained by the sample generating submodule according to a first preset proportion, and selecting a negative sample from a negative sample set obtained by the sample generating submodule to obtain a preset number of training samples;
correspondingly, the dividing module 203 is specifically configured to: and dividing the training samples of the preset number obtained by the selection sub-module into a training set and a test set according to a second preset proportion.
Wherein, the first preset proportion is preferably 1: 1; a second preset ratio, preferably 7: 3.
according to an embodiment of the present invention, the training module 204 specifically includes: a training submodule and an evaluation submodule, wherein:
the training submodule is used for training the training set divided by the dividing module 203 according to a random forest algorithm to obtain a prediction model;
in this embodiment, the training submodule is specifically configured to: and calling a random forest algorithm in the scimit-leann machine learning library by using a Python tool to train the training set divided by the dividing module 203 to obtain a prediction model.
The evaluation submodule is used for evaluating the prediction model obtained by the training submodule according to the working characteristic curve of the subject and adjusting the model parameters, and when the first area under the working characteristic curve of the subject meets a preset threshold value, the corresponding prediction model is output; wherein the preset threshold is preferably 0.8;
the test module 205 is specifically configured to: calculating a second area of the prediction model output by the evaluation sub-module under the working characteristic curve of the subject corresponding to the test set divided by the dividing module 203, and triggering the prediction module 206 when the second area meets a preset threshold;
correspondingly, the prediction module 206 is specifically configured to: with the current time as the cutoff time, the operation day number change trend, the number of times of parking at the maintenance station and the average parking duration of each vehicle in the third preset time period are counted forwards, and the future scrapped vehicle and the scrapping probability are predicted by using the prediction model tested by the testing module 205.
According to the method, the generation of the prediction model is carried out on the basis of historical track data of each vehicle in a certain period, the length and the width of the data are greatly increased, and the finally obtained prediction model and the prediction result are more scientific and reliable; meanwhile, the method is combined with machine learning, the operation rule of the vehicle before scrapping is found out according to a random forest algorithm, so that a prediction model for predicting scrapped vehicles and scrapping probability is obtained, and the problem of hysteresis in the prior art is solved; for manufacturers of all vehicles, marketing strategies can be determined more accurately.
The above description is only for the preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are included in the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the appended claims.
Claims (6)
1. A method of predicting a scrapped vehicle, comprising:
step S1: extracting characteristic information of each vehicle from the track data of each vehicle in a first preset time period;
step S2: generating training samples according to the characteristic information of each vehicle and a preset scrapped vehicle standard, and dividing the training samples into a training set and a test set;
step S3: training a prediction model according to the training set, and testing the prediction model by adopting the test set;
step S4: predicting future scrapped vehicles by using the tested prediction model;
the trajectory data includes: a series of position data and driving speed;
the step S1 specifically includes:
dividing the first preset time period into M second preset time periods, and taking the first n second preset time periods as third preset time periods according to the time sequence, wherein n is less than M;
according to the position data and the running speed contained in the track data of each vehicle, counting the number of operation days, the number of parking times of the maintenance station and the parking duration of each vehicle in each second preset time period;
determining the change trend of the number of operation days, the number of times of parking at the maintenance station and the average parking duration of each vehicle in the third preset time period according to the number of operation days, the number of times of parking at the maintenance station and the parking duration of each vehicle in each second preset time period;
determining the shutdown condition of each vehicle in each second preset time period according to the operation days of each vehicle in each second preset time period and the preset shutdown standard;
the step S3 specifically includes:
step S3-1: training the training set according to a random forest algorithm to obtain a prediction model;
step S3-2: evaluating the prediction model according to the working characteristic curve of the subject and adjusting the model parameters, and outputting the corresponding prediction model when the first area under the working characteristic curve of the subject meets a preset threshold value;
step S3-3: calculating a second area of the output prediction model under a corresponding subject work characteristic curve on the test set, and executing step S4 when the second area meets the preset threshold; otherwise, returning to the step S3-1;
the step S4 specifically includes: and counting the change trend of the operation days, the parking times of the maintenance station and the average parking duration of each vehicle in the third preset time period forwards by taking the current time as the cutoff time, and predicting the future scrapped vehicles and the scrapping probability by using the tested prediction model.
2. The method according to claim 1, wherein before the step S1, the method further comprises: preprocessing the track data of each vehicle in a first preset time period;
the step S1 specifically includes: and extracting characteristic information of each vehicle from the preprocessed track data.
3. The method according to claim 1, wherein the step S2 specifically includes:
step S2-1: determining scrapped vehicles and running vehicles in the vehicles according to the shutdown conditions of the vehicles in second preset time periods and preset scrapped vehicle standards;
step S2-2: establishing a corresponding relation between the operation day number change trend, the number of times of parking at a maintenance station and the average parking duration of each vehicle in the scrapped vehicles in the third preset time period and the corresponding vehicle, and using the corresponding relation as a positive sample to obtain a positive sample set; establishing a corresponding relation between the operation day number change trend, the number of times of parking at the maintenance station and the average parking duration of each vehicle in the running vehicles in the third preset time period and the corresponding vehicle, and using the corresponding relation as a negative sample to obtain a negative sample set;
step S2-3: selecting a positive sample from the positive sample set according to a first preset proportion, and selecting a negative sample from the negative sample set to obtain a preset number of training samples;
step S2-4: and dividing the training samples of the preset number into a training set and a test set according to a second preset proportion.
4. A system for predicting a scrapped vehicle, comprising:
the extraction module is used for extracting the characteristic information of each vehicle from the track data of each vehicle in a first preset time period;
the generating module is used for generating training samples according to the characteristic information of each vehicle extracted by the extracting module and a preset scrapped vehicle standard;
the dividing module is used for dividing the training samples generated by the generating module into a training set and a test set;
the training module is used for training a prediction model according to the training set divided by the dividing module;
the testing module is used for testing the prediction model trained by the training module by adopting the testing set divided by the dividing module;
the prediction module is used for predicting a future scrapped vehicle by using the prediction model tested by the test module;
the trajectory data includes: a series of position data and driving speed;
the extraction module is specifically configured to:
dividing the first preset time period into M second preset time periods, and taking the first n second preset time periods as third preset time periods according to the time sequence, wherein n is less than M;
according to the position data and the running speed contained in the track data of each vehicle, counting the number of operation days, the number of parking times of the maintenance station and the parking duration of each vehicle in each second preset time period;
determining the change trend of the number of operation days, the number of times of parking at the maintenance station and the average parking duration of each vehicle in the third preset time period according to the number of operation days, the number of times of parking at the maintenance station and the parking duration of each vehicle in each second preset time period;
determining the shutdown condition of each vehicle in each second preset time period according to the operation days of each vehicle in each second preset time period and the preset shutdown standard;
the training module specifically comprises: a training submodule and an evaluation submodule;
the training submodule is used for training the training set divided by the dividing module according to a random forest algorithm to obtain a prediction model;
the evaluation submodule is used for evaluating the prediction model obtained by the training submodule according to the working characteristic curve of the subject and adjusting the model parameters, and when the first area under the working characteristic curve of the subject meets a preset threshold value, the corresponding prediction model is output;
the test module is specifically configured to: calculating a second area of the prediction model output by the evaluation sub-module under a corresponding test subject working characteristic curve on the test set divided by the dividing module, and triggering the prediction module when the second area meets the preset threshold;
the prediction module is specifically configured to: and counting the change trend of the operation days, the parking times of the maintenance station and the average parking duration of each vehicle in the third preset time period forwards by taking the current time as a cut-off time, and predicting the future scrapped vehicle and the scrapping probability by using the prediction model tested by the test module.
5. The system of claim 4, further comprising: a preprocessing module;
the preprocessing module is used for preprocessing the track data of each vehicle in a first preset time period;
the extraction module is specifically configured to: and extracting the characteristic information of each vehicle from the track data preprocessed by the preprocessing module.
6. The system of claim 5, wherein the generating module specifically comprises: determining a submodule, a sample generation submodule and a selection submodule;
the determining submodule is used for determining the scrapped vehicles and the running vehicles in the vehicles according to the shutdown conditions of the vehicles in the second preset time period and the preset scrapped vehicle standards extracted by the extracting module;
the sample generation submodule is used for establishing a corresponding relation between the operation day number change trend, the number of times of parking at the maintenance station and the average parking duration of each vehicle in the scrapped vehicles determined by the determination submodule in the third preset time period and the corresponding vehicle, and using the corresponding relation as a positive sample to obtain a positive sample set; establishing a corresponding relation between the operation day number change trend, the number of times of parking at the maintenance station and the average parking duration of each vehicle in the running vehicles in the third preset time period, which are determined by the determination sub-module, and the corresponding vehicle, and taking the corresponding relation as a negative sample to obtain a negative sample set;
the selecting submodule is used for selecting a positive sample from the positive sample set obtained by the sample generating submodule according to a first preset proportion, and selecting a negative sample from the negative sample set obtained by the sample generating submodule to obtain a preset number of training samples;
the dividing module is specifically configured to: and dividing the training samples of the preset number obtained by the selection submodule into a training set and a test set according to a second preset proportion.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711445897.9A CN108053075B (en) | 2017-12-27 | 2017-12-27 | Scrapped vehicle prediction method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711445897.9A CN108053075B (en) | 2017-12-27 | 2017-12-27 | Scrapped vehicle prediction method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108053075A CN108053075A (en) | 2018-05-18 |
| CN108053075B true CN108053075B (en) | 2021-03-26 |
Family
ID=62128185
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711445897.9A Active CN108053075B (en) | 2017-12-27 | 2017-12-27 | Scrapped vehicle prediction method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108053075B (en) |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109102149A (en) * | 2018-06-06 | 2018-12-28 | 北京市燃气集团有限责任公司 | A kind of prediction technique of city gas buried pipeline third party breakage in installation risk |
| CN108810526B (en) * | 2018-06-12 | 2020-04-10 | 广东惠禾科技发展有限公司 | Camera fault intelligent prediction method and device and camera management server |
| CN109345136A (en) * | 2018-10-22 | 2019-02-15 | 中山大学 | Vehicle survival curve model optimization method based on China's motor vehicle resignation system |
| CN109635965A (en) * | 2018-12-24 | 2019-04-16 | 成都四方伟业软件股份有限公司 | Bus scraps decision-making technique, device and readable storage medium storing program for executing |
| CN111861073B (en) * | 2019-04-30 | 2024-05-03 | 北京嘀嘀无限科技发展有限公司 | Operation policy prediction method and device, electronic equipment and readable storage medium |
| CN111144485B (en) * | 2019-12-26 | 2023-07-21 | 北京中交兴路车联网科技有限公司 | Vehicle accident judging method and system based on xgboost classification algorithm |
| CN111860954B (en) * | 2020-06-18 | 2024-07-05 | 上海钧正网络科技有限公司 | Vehicle loss-of-link prediction method, device, computer equipment and storage medium |
| CN112529104B (en) * | 2020-12-23 | 2024-06-18 | 东软睿驰汽车技术（沈阳）有限公司 | Vehicle fault prediction model generation method, fault prediction method and device |
| CN113570469B (en) * | 2021-07-27 | 2024-05-28 | 上海交通大学 | Intelligent vehicle change prediction method for vehicle insurance user |
| CN114912675A (en) * | 2022-05-06 | 2022-08-16 | 杉数科技(北京)有限公司 | Prediction method and device for scrappage of aero-engine blade |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105389974A (en) * | 2015-11-19 | 2016-03-09 | 深圳市赛格导航科技股份有限公司 | Vehicle tracking method and system based on vehicle historical driving data |
| CN105868865A (en) * | 2016-04-13 | 2016-08-17 | 华北电力大学 | Electric vehicle parc prediction method based on multivariate linear regression method and proportional substitution method |
| CN106127312A (en) * | 2016-06-21 | 2016-11-16 | 中车青岛四方机车车辆股份有限公司 | The life cycle management history information management method of EMUs and system |
| CN107025345A (en) * | 2017-03-31 | 2017-08-08 | 柳州顺联科技有限公司 | A kind of Forecasting Methodology of engineering machinery vehicle failure time |
-
2017
- 2017-12-27 CN CN201711445897.9A patent/CN108053075B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105389974A (en) * | 2015-11-19 | 2016-03-09 | 深圳市赛格导航科技股份有限公司 | Vehicle tracking method and system based on vehicle historical driving data |
| CN105868865A (en) * | 2016-04-13 | 2016-08-17 | 华北电力大学 | Electric vehicle parc prediction method based on multivariate linear regression method and proportional substitution method |
| CN106127312A (en) * | 2016-06-21 | 2016-11-16 | 中车青岛四方机车车辆股份有限公司 | The life cycle management history information management method of EMUs and system |
| CN107025345A (en) * | 2017-03-31 | 2017-08-08 | 柳州顺联科技有限公司 | A kind of Forecasting Methodology of engineering machinery vehicle failure time |
Non-Patent Citations (1)
| Title |
|---|
| 基于灰色***理论的汽车报废量预测;赵鹏等;《工业技术经济》;20110430;第31页摘要-第35页第3节 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108053075A (en) | 2018-05-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108053075B (en) | Scrapped vehicle prediction method and system | |
| CN104272328B (en) | For the method for the forecast consumption value for determining motor vehicle | |
| CN109767597B (en) | Vehicle accident early warning method and system | |
| CN101950477B (en) | Method and device for processing traffic information | |
| CN106650157B (en) | Method, device and system for estimating fault occurrence probability of vehicle parts | |
| CN108195396B (en) | Method and system for detecting validity of new energy automobile data | |
| CN106406287A (en) | Method and system for vehicle safety state monitoring and early warning | |
| CN104809878A (en) | Method for detecting abnormal condition of urban road traffic by utilizing GPS (Global Positioning System) data of public buses | |
| CN112036581B (en) | Performance detection method and device for vehicle air conditioning system, storage medium and terminal | |
| CN111144485A (en) | Vehicle accident judgment method and system based on xgboost classification algorithm | |
| CN103606247B (en) | Traffic early-warning method and system by means of vehicle conditions and driver physiological parameters | |
| CN117392855B (en) | Vehicle overrun overload probability identification method and system based on satellite positioning data | |
| CN103714695A (en) | Method for analyzing vehicle operation state based on vehicle running data | |
| CN102622883A (en) | Method and device for judging release of traffic incident | |
| CN115758227A (en) | High-energy-consumption vehicle identification and reason diagnosis method | |
| CN116380496B (en) | Automobile door fatigue endurance test method, system and medium | |
| CN109840566A (en) | Trip mode identification method and device | |
| CN114676870A (en) | Vehicle speed prediction method, vehicle speed prediction model training method and related device | |
| CN115131895B (en) | Method and device for acquiring vehicle mileage data and electronic equipment | |
| CN110291470B (en) | Method for determining support points for a test plan | |
| CN108197986B (en) | Vehicle prediction method and system | |
| CN112699490B (en) | Vehicle maintenance result verification method and device | |
| Khemri et al. | Developing a Real-World, Second-by-Second Driving Cycle Database through Public Vehicle Trip Surveys | |
| CN111539686B (en) | Transportation management monitoring method based on vehicle running record visualization | |
| Zhang et al. | Pass-by noise modelling applying machine learning |
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 |