Prediction method of remaining urea usage time based on engine data

Abstract

The present invention relates to a method for predicting the remaining use time of urea based on engine data. Different from the previous method of calculating by ECU, a big data platform is adopted to realize data storage, calculation and display, which is more conducive to the monitoring and management of the whole life cycle of the vehicle and digital empowerment. The urea adding time point is accurately identified by an algorithm, so as to select the latest period of normal use data of urea, and the urea consumption rate is calculated by the least square method, so as to effectively reduce the deviation of the calculation result caused by the acquisition error, and the remaining use time of urea can be dynamically predicted, which is closer to the actual situation, and is convenient for users to reasonably arrange the time of adding urea, so as to avoid excessive emission and unnecessary frequent addition of urea due to untimely addition of urea. The method has few prerequisites, and only requires the liquid level signal of the urea tank, and has low requirements on the signal quality.

Description

Technical Field

[0001] The invention relates to a vehicle safety technology, and in particular to a method for predicting the remaining use time of urea based on engine data. Background Art

[0002] The traditional technology for calculating the remaining cruising range of urea mainly relies on the calculation of the average urea consumption per 100 kilometers and the remaining urea to obtain the cruising range of the remaining urea. The disadvantage of this technical solution is that it only considers the urea consumption per 100 kilometers, that is, it predicts the cruising range of the remaining urea based on the previous user's driving habits. If the user's vehicle load changes drastically or the driving conditions change drastically, the remaining cruising range predicted by this technical solution will have a large deviation, resulting in the failure to remind the user to add urea in time, causing inconvenience to the user.

[0003] The existing technology stores and calculates data through ECU. Historical data cannot be stored or shared, and it is impossible to further empower fleet management through data.

[0004] The existing technology has many preconditions. It is necessary to know the maximum urea amount of the vehicle and preset the urea amount. Different models of vehicles need to be configured with different parameters, resulting in low scalability and coverage of the method.

[0005] The existing technology has high requirements for the quality of the urea liquid level sensor signal, but the system error contained in the signal is inevitable, so the applicability of the existing technology is relatively low. Summary of the invention

[0006] In order to solve the problem that the remaining urea calculation method is unreasonable, resulting in a large deviation in the predicted cruising range of the remaining urea, a method for predicting the remaining urea usage time based on engine data is proposed. The real-time engine operation data is obtained through intelligent sensing, on-board terminal TBOX and 4G communication. Based on technologies such as intelligent networking, big data and machine learning, the remaining urea usage time prediction is realized, which is convenient for the driver to control the current urea usage and reasonably arrange the time for adding urea to prevent excessive emissions and unnecessary frequent urea addition due to untimely urea addition.

[0007] The technical scheme of the present invention is: a method for predicting the remaining use time of urea based on engine data, using a vehicle-mounted terminal to collect the urea tank level, the NE activation signal corresponding to the crankshaft rotation signal, and the actual urea injection amount in real time, and extracting the maximum urea tank level per hour as the urea tank level characteristic value of the hour after the cloud platform data is cleaned; firstly, the time point of adding urea is determined by identifying the time when the urea level characteristic value suddenly changes upward, extracting the subsequent data of the determined urea adding time, selecting the most recent data segment without urea addition from the extracted data, using the least squares method to fit the urea consumption time-urea tank level characteristic value, and obtaining the slope of the straight line, which is the urea consumption rate per hour, and then dividing the moving average of the urea tank level in the current period by the slope of the straight line to obtain the remaining use time of urea, and recording the remaining use time of urea, the current urea tank level, and the urea consumption rate at this time; finally, triggering corresponding alarm push according to different thresholds corresponding to the remaining use time of urea.

[0008] Further, the specific method for identifying the time when the urea liquid level characteristic value suddenly changes upward includes the following steps:

[0009] 1) Data collection: Count the number of data collected every hour, record it as the number of data per hour, extract the maximum urea tank liquid level per hour as the urea tank liquid level characteristic value of that hour, and store the urea tank liquid level characteristic value and the number of data per hour in the characteristic value intermediate table;

[0010] 2) Data filtering: read the intermediate table of vehicle characteristic values, sort them by time, and delete the abnormal points. The abnormal points are the points where the characteristic value of the urea tank level is 20 points lower than before or after, or the points where the characteristic value of the urea tank level is 5 points higher than before;

[0011] 3) Data smoothing: The urea tank liquid level characteristic value is averaged and smoothed with a window length of 2 to obtain a smoothed urea tank liquid level characteristic value, thereby reducing the impact of the mutation point on the calculation of the remaining use time of urea;

[0012] 4) Data screening, extraction of urea addition time:

[0013] Arrange the processed data in chronological order, and add a serial number to the sorted data. The serial number indicates the running hour to which the corresponding data belongs.

[0014] Filter the data whose "smoothed urea tank liquid level characteristic value" is 5 higher than the previous one and not lower than the next one, or the data whose current urea tank liquid level characteristic value is 10 higher than the previous one, or the data whose current urea tank liquid level characteristic value is more than 20 lower than the previous one, or the data whose natural time difference between the current urea tank liquid level characteristic value point and the previous one is more than 7 days. If any of the above appears in adjacent sequence numbers, the adjacent sequence numbers retain the time in the smaller sequence number as the urea tank liquid level mutation time point. The obtained urea tank liquid level mutation time point is the extracted urea addition time, which is used for data segmentation.

[0015] Further, the remaining use time of urea is calculated as follows:

[0016] If the urea adding time is not extracted, all data are selected and saved; if the urea adding time is extracted, the data after the urea adding time is extracted as the data for this segmentation. The 50 most recent data for this segmentation are used to fit a straight line using the least squares method. The calculated straight line slope is multiplied by -1 to become a positive number. The remaining urea usage time = the latest "smoothed urea tank level" / -1 multiplied by the slope.

[0017] Further, the method of pushing the alarm is as follows: taking the urea remaining usage time result table, and pushing the alarm in segmented time:

[0018] If the remaining usage time is between 20 and 30 hours for three consecutive times, and the third time is less than the first time, an alarm will be pushed: the remaining usage time of urea is less than 30 hours;

[0019] If the remaining usage time is between 10 and 20 hours for three consecutive times, and the third time is less than the first time, an alarm will be pushed: the remaining usage time of urea is less than 20 hours;

[0020] The remaining usage time is less than 10 hours, and the alarm is pushed: the remaining usage time of urea is less than 10 hours.

[0021] An application of a urea remaining usage time prediction method based on engine data associates urea usage time with corresponding mileage and calculates statistics. The urea remaining usage time is predicted by the urea remaining usage time prediction method based on engine data, and the predicted urea remaining usage mileage is obtained accordingly.

[0022] The beneficial effects of the present invention are as follows: the method for predicting the remaining use time of urea based on engine data is different from the previous method of calculating by ECU. It adopts a big data platform to realize data storage, calculation and display, which is more conducive to the monitoring and management of the whole life cycle of the vehicle and digital empowerment; the algorithm accurately identifies the time point of adding urea, thereby selecting the latest data of normal use of urea (which can better reflect the urea use status of the current vehicle), and uses the least squares method to calculate the urea consumption rate, which effectively reduces the deviation of the calculation result caused by the collection error, and can dynamically predict the remaining use time of urea, which is closer to the actual situation and convenient for use. Users can arrange the time of adding urea reasonably to avoid excessive emissions and unnecessary frequent addition of urea due to untimely addition of urea; there are few prerequisites, only the urea tank liquid level signal is required, and the signal quality requirements are low. In this method, the fluctuation of signal acquisition can be reduced through data cleaning and aggregation algorithms, which greatly improves the universality; through the data of nearly a thousand vehicles from 20220401 to 20220701, a total of 55 vehicles with less than 30 hours of remaining urea were identified, of which 51 vehicles added urea shortly after the alarm, with an accuracy rate of 51 / 55=92.72%. The accuracy can be further improved through training and optimization of large amounts of data in the later stage. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 A data realization diagram of a method for predicting the remaining use time of urea based on engine data of the present invention;

[0024] Figure 2a It is a flow chart of extracting the intermediate value of the characteristic value in the method of the present invention;

[0025] Figure 2b A flow chart of calculating the remaining use time of a vehicle in the method of the present invention;

[0026] Figure 2c This is a flow chart of insufficient urea remaining time alarm in the method of the present invention;

[0027] Figure 3 This is a schematic diagram of a first embodiment of the method of the present invention;

[0028] Figure 4 This is a schematic diagram of Embodiment 2 of the method of the present invention;

[0029] Figure 5 This is a schematic diagram of Embodiment 3 of the method of the present invention. DETAILED DESCRIPTION

[0030] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. This embodiment is implemented based on the technical solution of the present invention, and provides a detailed implementation method and specific operation process, but the protection scope of the present invention is not limited to the following embodiments.

[0031] like Figure 1 The data implementation diagram of the urea remaining use time prediction method based on engine data is shown. The vehicle terminal TBOX is used to collect the urea tank level, the NE activation signal corresponding to the crankshaft rotation signal, and the actual urea injection amount in real time. After the cloud platform data is cleaned, the maximum urea tank level per hour is extracted as the urea tank level characteristic value of that hour. First, the time when the urea level characteristic value suddenly changes upward (the time point of adding urea) is identified, and the subsequent data of the determined urea adding time is extracted. The most recent data segment without urea addition is taken from the extracted data. The least squares method is used to fit the urea consumption time-urea tank level characteristic value to obtain the straight line slope (that is, the hourly urea consumption rate). Then, the moving average of the urea tank level in the current period is divided by the straight line slope to obtain the urea remaining use time. The urea remaining use time, remaining mileage, current urea tank level, and urea consumption rate at this time are recorded. Finally, the corresponding alarm push is triggered according to the different thresholds corresponding to the urea remaining use time.

[0032] Input & output signals, TBOX needs to collect three input signals and upload them. After cloud judgment, it outputs the corresponding judgment results. The signal list is as follows:

[0033]

[0034] The urea remaining use time prediction method based on engine data includes three steps: extracting the intermediate value of the characteristic value, calculating the remaining use time of the vehicle, and warning of insufficient remaining time of urea. Figure 2a , 2b , as shown in 2c.

[0035] Step 1, extraction of intermediate characteristic value: collect the crankshaft rotation signal (NE activation) as 1 in real time, collect the urea tank level and the actual urea injection amount in real time, count the maximum urea tank level per hour, record it as the urea tank level statistics, count the number of data collected per hour, record it as the number of data per hour. Extract the maximum urea tank level per hour as the characteristic value of the urea tank level for that hour, and store it in the intermediate characteristic value table together with the number of data per hour.

[0036] Step 2: Calculate the remaining usage time of urea:

[0037] 2.1、Data filtering: Read the intermediate table of vehicle characteristic values, sort them by time, and delete the abnormal points collected, because in a period of time when there is no sudden upward change in the characteristic value of urea liquid level, the urea liquid level will only move downward. The collected urea tank liquid level may not be the actual liquid level value due to vehicle vibration or other conditions. Delete the points where the characteristic value of urea tank liquid level is 20 lower than before and after, and delete the points where the characteristic value of urea tank liquid level is 5 higher than before and after.

[0038] 2.2. Data smoothing: The characteristic value of the urea tank liquid level is averaged and smoothed with a window length of 2 to obtain the smoothed urea tank liquid level, thereby reducing the impact of the mutation point on the calculation of the remaining use time of urea.

[0039] 2.3. Data screening, extraction and urea addition time:

[0040] Arrange the filtered data in chronological order, and add serial numbers to the sorted data. The serial numbers indicate the running hours to which the corresponding data belongs.

[0041] Filter the data whose "smoothed urea tank liquid level characteristic value" is 5 higher than the previous one and not lower than the next point, or the current urea tank liquid level characteristic value is 10 higher than the previous point, or the current urea tank liquid level characteristic value is more than 20 lower than the previous point, or the natural time difference between the current urea tank liquid level characteristic value point and the previous point is more than 7 days. If there are adjacent sequence numbers with one of the above, the adjacent sequence numbers retain the time in the smaller sequence number as the urea tank liquid level mutation time point. The obtained urea tank liquid level mutation time point is the extracted urea addition time, which is used for data segmentation.

[0042] 2.4. Whether the urea addition time can be extracted.

[0043] 2.5. Data segmentation: If the urea adding time is not extracted, all data are selected for saving; if the urea adding time is extracted, the data after the urea adding time is extracted as the segmentation data this time, and the data with a sequence number greater than the maximum sequence number of the last data segmentation is selected from the segmentation data.

[0044] 2.6. Draw the slope of the straight line from the split data and calculate the remaining use time of urea: Determine whether the number of selected data is greater than 10. If not, skip step 2 and return to step 1. If greater than 10, the urea level change rate (% / h) = the 50 most recent data from the split data, fit the straight line using the least squares method, and calculate the slope of the straight line (*-1 becomes a positive number). The remaining use time of urea = the latest "smoothed urea tank level" / slope * (-1).

[0045] 2.7. The results are stored in the urea remaining usage time result table.

[0046] Step 3: Read the urea remaining usage time result table and push segmented time alarms:

[0047] If the remaining usage time is 20-30h for three consecutive times and the third time is less than the first time, an alarm will be pushed: the remaining usage time of urea is less than 30h; if the remaining usage time is 10-20h for three consecutive times and the third time is less than the first time, an alarm will be pushed: the remaining usage time of urea is less than 20h; if the remaining usage time is less than 10h, an alarm will be pushed: the remaining usage time of urea is less than 10h.

[0048] Embodiment 1

[0049] TBOX collects engine operation data and sends it to the big data platform. The big data platform calculates and identifies the mutation point of the urea tank level (that is, the time point for adding urea), intercepts the latest data segment without urea addition (reduces the impact of historical data), and takes nearly 50 data points (more in line with the recent vehicle urea usage) to fit the time-urea tank level characteristic value using the least squares method to obtain the slope of the straight line (that is, the hourly urea consumption rate). The current urea tank level moving average (moving average can eliminate the impact of system collection data errors) is divided by the slope to obtain the remaining urea usage time. A record is generated when the conditions are met.

[0050] like Figure 3 The point on the warning line is the remaining urea usage time for that hour, and the point below the warning line is the remaining urea usage time that is actually less than 30 hours. The predicted remaining urea usage time is dynamically changed according to the changes in the urea tank liquid level in recent period of time, which is more in line with the actual situation.

[0051] Embodiment 2

[0052] TBOX collects engine operation data and sends it to the big data platform. The big data platform calculates and identifies the mutation point of the urea tank level (that is, the time point for adding urea), intercepts the latest data segment without urea addition (reduces the impact of historical data), and takes nearly 50 data points (more in line with the recent vehicle urea usage) to fit the time-urea tank level characteristic value using the least squares method to obtain the slope of the straight line (that is, the hourly urea consumption rate). The current urea tank level moving average (moving average can eliminate the impact of system collection data errors) is divided by the slope to obtain the remaining urea usage time. A record is generated when the conditions are met.

[0053] like Figure 4 The slopes of the three urea decline sections are obviously different. This scheme only uses the latest section of data to calculate the urea consumption rate, eliminating the impact of historical data.

[0054] Embodiment 3

[0055] TBOX collects engine operation data and sends it to the big data platform. The big data platform calculates and identifies the mutation point of the urea tank level (that is, the time point for adding urea), intercepts the latest data segment without urea addition (reduces the impact of historical data), and takes nearly 50 data points (more in line with the recent vehicle urea usage) to fit the time-urea tank level characteristic value using the least squares method to obtain the slope of the straight line (that is, the hourly urea consumption rate). The current urea tank level moving average (moving average can eliminate the impact of system collection data errors) is divided by the slope to obtain the remaining urea usage time. A record is generated when the conditions are met.

[0056] like Figure 5 As shown in the figure, the dark dots represent the urea tank liquid level when the remaining usage time of urea is less than 30h (i.e. the Y-axis on the left side of the figure below), and the light dots represent the corresponding remaining time (i.e. the Y-axis on the right side of the figure below). It can be seen that the urea tank liquid level is significantly increased after the alarm is triggered (the driver added urea), indicating that the model is consistent with the actual situation.

[0057] The urea remaining usage time prediction method can also be used to predict the remaining usage mileage of urea. As long as the urea usage time is statistically associated with the corresponding mileage, when the urea remaining usage time is predicted, the corresponding predicted urea remaining usage mileage can be obtained.

[0058] The above-mentioned embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. A method for predicting the remaining use time of urea based on engine data, characterized in that: The vehicle-mounted terminal is used to collect the urea tank level, the NE activation signal corresponding to the crankshaft rotation signal, and the actual urea injection amount in real time. After the cloud platform data is cleaned, the maximum urea tank level per hour is extracted as the urea tank level characteristic value of that hour. First, the time point of adding urea is determined by identifying the time when the characteristic value of the urea liquid level suddenly changes upward, and the subsequent data of the determined urea adding time is extracted. The most recent data segment without urea addition is selected from the extracted data, and the least squares method is used to fit the urea consumption time-urea tank liquid level characteristic value to obtain the slope of the straight line. The slope of the straight line is the urea consumption rate per hour, and then the moving average value of the urea tank liquid level in the current period is divided by the slope of the straight line to obtain the remaining urea use time, and the remaining urea use time, current urea tank liquid level, and urea consumption rate at this time are recorded; finally, the corresponding alarm push is triggered according to different thresholds corresponding to the remaining urea use time; The specific method for identifying the time of the upward mutation of the urea liquid level characteristic value comprises the following steps: 1) Data collection: Count the number of data collected every hour, record it as the number of data per hour, extract the maximum urea tank liquid level per hour as the urea tank liquid level characteristic value of that hour, and store the urea tank liquid level characteristic value and the number of data per hour in the characteristic value intermediate table; 2) Data filtering: read the intermediate table of vehicle characteristic values, sort them by time, and delete the abnormal points. The abnormal points are the points where the characteristic value of the urea tank level is 20 points lower than before or after, or the points where the characteristic value of the urea tank level is 5 points higher than before; 3) Data smoothing: The urea tank liquid level characteristic value is averaged and smoothed with a window length of 2 to obtain a smoothed urea tank liquid level characteristic value, thereby reducing the impact of the mutation point on the calculation of the remaining use time of urea; 4) Data screening, extraction of urea addition time: Arrange the processed data in chronological order, and add a serial number to the sorted data. The serial number indicates the running hour to which the corresponding data belongs. Filter the data whose smoothed urea tank liquid level characteristic value is 5 higher than the previous one and not lower than the next one, or the data whose current urea tank liquid level characteristic value is 10 higher than the previous one, or the data whose current urea tank liquid level characteristic value is more than 20 lower than the previous one, or the data whose natural time difference between the current urea tank liquid level characteristic value point and the previous one is more than 7 days. If any of the above occurs in adjacent sequence numbers, the adjacent sequence numbers retain the time in the smaller sequence number as the urea tank liquid level mutation time point. The obtained urea tank liquid level mutation time point is the extracted urea addition time, which is used for data segmentation.

2. The method for predicting the remaining use time of urea based on engine data according to claim 1, characterized in that: The calculation method of the remaining use time of urea is: If the urea adding time is not extracted, all data are selected and saved; if the urea adding time is extracted, the data after the urea adding time is extracted as the data for this segmentation. The 50 most recent data for this segmentation are used to fit a straight line using the least squares method. The calculated straight line slope is multiplied by -1 to become a positive number. The remaining urea usage time = the latest smoothed urea tank level / -1 multiplied by the slope.

3. The method for predicting the remaining use time of urea based on engine data according to claim 2, characterized in that: The method of pushing the alarm is as follows: taking the urea remaining usage time result table, and pushing the alarm in segmented time: If the remaining usage time is between 20 and 30 hours for three consecutive times, and the third time is less than the first time, an alarm will be pushed: the remaining usage time of urea is less than 30 hours; If the remaining usage time is between 10 and 20 hours for three consecutive times, and the third time is less than the first time, an alarm will be pushed: the remaining usage time of urea is less than 20 hours; The remaining usage time is less than 10 hours, and the alarm is pushed: the remaining usage time of urea is less than 10 hours.

Images