Equipment working hour statistical method and system based on histogram distance measurement

Abstract

The invention relates to the technical field of man-hour statistics, in particular to an equipment man-hour statistics method and system based on histogram distance measurement. The method comprises the following steps: acquiring a mechanical equipment vibration signal as idle speed data; generating an amplitude histogram according to the obtained mechanical equipment vibration signal; storing the generated amplitude histogram as prior distribution of idle speed data; and carrying out state identification on the vibration signal by using the stored amplitude histogram. The abnormal points in the vibration signals are specially processed, only the idle speed data need to be collected through the method, the method is simpler, and due to the fact that manual labeling is not needed, a large amount of labeling cost can be saved.

Description

Technical Field

[0001] The present invention relates to the technical field of man-hour statistics, and in particular to a method and system for equipment man-hour statistics based on histogram distance measurement. Background Art

[0002] The infrastructure industry relies heavily on construction machinery. The actual working conditions of construction machinery on site are currently mainly based on shifts, which are generally settled in half-day units. This method is prone to false reporting of working hours and difficulties in supervision, which damages the economic interests of enterprises.

[0003] According to the vibration signals generated by the construction machinery during operation, it can be judged whether the machinery is in the idle state or the working state. By accumulating the vibration signal time in the working state, the actual working time of the construction machinery can be obtained. This is also the mainstream method currently used. After collecting a large number of vibration signals, the existing methods need to manually mark the operating status labels in the corresponding time period in order to extract features, use supervised learning methods to identify the idle and working states, and then count the equipment working hours based on this. This type of method requires a large amount of data collection and labeling in the early stage, resulting in high labor costs. At the same time, the trained model has a strong correlation with the data, and the model generalization ability is poor. After replacing the equipment, it is necessary to re-collect and manually label again before training the model. For a large number of engineering equipment, it is almost impossible to train different models for each device. Summary of the invention

[0004] In order to solve the above-mentioned problems, the present invention provides a method and system for counting equipment working hours based on histogram distance measurement.

[0005] In a first aspect, the present invention provides a method for counting equipment working hours based on histogram distance measurement, which adopts the following technical solution: A method for counting equipment working hours based on histogram distance measurement, comprising: Obtain mechanical equipment vibration signals as idle data; Generate an amplitude histogram according to the acquired mechanical equipment vibration signal; The generated amplitude histogram is stored in the database as the prior distribution of the idle data; The stored amplitude histogram is used to identify the state of the vibration signal.

[0006] Furthermore, the obtaining of the vibration signal of the mechanical equipment includes collecting the original vibration signal of the mechanical equipment through a vibration signal sensor, with a collection frequency of 300 Hz, a sampling time of 1 s, and a sampling interval of 1 minute. The original vibration signal is sent to the cloud server through a gateway device using the MQTT protocol.

[0007] Furthermore, the amplitude histogram is generated based on the acquired mechanical equipment vibration signal, including obtaining quartiles by arranging the data of the vibration signal, counting upper and lower abnormal points according to the obtained quartiles, and generating a normal amplitude histogram for the data with amplitudes between the upper and lower abnormal points through counting the upper and lower abnormal points.

[0008] Furthermore, the amplitude histogram is generated according to the acquired mechanical equipment vibration signal, and also includes histogram splicing of the normal amplitude histogram, wherein the lower abnormal point count value is placed at the front end of the histogram, and the upper abnormal point count value is placed at the end of the histogram to obtain a spliced ​​amplitude histogram.

[0009] Furthermore, the generating of the amplitude histogram according to the acquired mechanical equipment vibration signal also includes normalizing the spliced ​​amplitude histogram, wherein the histogram count of the spliced ​​amplitude histogram is divided by the total number of points to obtain the normalized amplitude histogram.

[0010] Furthermore, the generated amplitude histogram is stored in the library as the prior distribution of the idle data, including fixing the sensor on engineering machinery of different manufacturers and sizes and collecting a large amount of idle data, generating a normalized histogram of the sampling each time, storing all the normalized histograms in the library as the prior distribution of the idle data, assuming that a total of w times, the size of the idle bank H is ,in b is the number of groups in the histogram.

[0011] Furthermore, the state identification of the vibration signal is performed using the stored amplitude histogram, including calculating the Euclidean distance between the vibration signal and all the data in the idle library according to the amplitude histogram, thereby obtaining the average Euclidean distance between the vibration signal and all the data in the idle library, and obtaining the state identification result of the vibration signal by setting a threshold.

[0012] In a second aspect, a device working time statistics system based on histogram distance measurement includes: The data acquisition module is configured to acquire a vibration signal of the mechanical equipment as idle speed data; A histogram module is configured to generate an amplitude histogram according to the acquired mechanical equipment vibration signal; The storage module is configured to store the generated amplitude histogram as a prior distribution of the idle data; The state recognition module is configured to use the stored amplitude histogram to perform state recognition of the vibration signal.

[0013] In a third aspect, the present invention provides a computer-readable storage medium storing a plurality of instructions, wherein the instructions are suitable for being loaded and executed by a processor of a terminal device, for example, a method for counting equipment working hours based on a histogram distance metric.

[0014] In a fourth aspect, the present invention provides a terminal device comprising a processor and a computer-readable storage medium, wherein the processor is used to implement various instructions; the computer-readable storage medium is used to store multiple instructions, and the instructions are suitable for being loaded by the processor and executing the equipment working time statistics method based on histogram distance measurement.

[0015] In summary, the present invention has the following beneficial technical effects: The present invention uses a method based on histogram distance measurement to perform working hours statistics of engineering machinery. By referring to the bag-of-words model in text classification, the vibration signal is divided into multiple segments with different numerical ranges to generate an amplitude histogram. Each segment of the histogram is equivalent to a "word" in the bag-of-words model, and the size of the histogram is equivalent to the "word frequency".

[0016] At the same time, special processing is performed on abnormal points in the vibration signal. Through this method, only idle data needs to be collected. The method is simpler and can save a lot of labeling costs because no manual labeling is required. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 is a schematic diagram of a method for counting equipment working hours based on histogram distance measurement according to Embodiment 1 of the present invention; Figure 2 is a schematic diagram of data collection in Example 1 of the present invention; Figure 3 is a normal amplitude histogram of the idle data of Embodiment 1 of the present invention; Figure 4 is a splicing amplitude histogram of the idle data of Example 1 of the present invention; Figure 5 is a normalized amplitude histogram of embodiment 1 of the present invention; Figure 6 This is an example diagram of local data of the idle library according to Embodiment 1 of the present invention; Figure 7 It is a schematic diagram of the recognition result of data in Example 1 of the present invention. DETAILED DESCRIPTION

[0018] The present invention is further described in detail below in conjunction with the accompanying drawings.

[0019] Example 1 The present invention uses a method based on histogram distance measurement to perform working hours statistics of engineering machinery. By referring to the bag-of-words model in text classification, the vibration signal is divided into multiple segments with different numerical ranges to generate an amplitude histogram. Each segment of the histogram is equivalent to a "word" in the bag-of-words model, and the size of the histogram is equivalent to the "word frequency". The amplitude histogram reflects the amplitude distribution of the vibration signal. In particular, the present invention performs special processing on abnormal points in the vibration signal, which will be described in detail in the subsequent steps.

[0020] The present invention first collects a large amount of idle data, generates an amplitude histogram of all idle data and stores it in a database, which is used as a priori for idle data. Next, the average Euclidean distance between the newly collected data and the idle database data is calculated. , set the distance threshold t Afterwards, when ≤ t , it is judged as idling, otherwise the running state is working. The overall process is as follows Figure 1 This method only needs to collect idle data, which is simpler and can save a lot of labeling costs because it does not require manual labeling.

[0021] Reference Figure 1 , a method for counting equipment working hours based on histogram distance measurement in this embodiment includes: Obtain mechanical equipment vibration signals as idle data; Generate an amplitude histogram according to the acquired mechanical equipment vibration signal; The generated amplitude histogram is stored in the database as the prior distribution of the idle data; The stored amplitude histogram is used to identify the state of the vibration signal.

[0022] Specifically: 1. Data Collection Use a vibration signal sensor to collect mechanical vibration signals with a sampling frequency of 300 Hz, a sampling time of 1 second, and a sampling interval of 1 minute. Figure 2 As shown: The original vibration signal is sent to the cloud server via the gateway device using the MQTT protocol. On the server side, 300 sample points are processed each time.

[0023] 2. Generate an amplitude histogram The amount of data processed each time is 300 vibration signal samples within 1 second.

[0024] (1) Calculate the upper and lower quartiles Assume that the number of data points is n , arrange all the values ​​from small to large and divide them into four equal parts, and the values ​​at the three dividing points are the quartiles.

[0025] set up and Respectively The integer and decimal parts of and Respectively The fractional part of the integer part of Q 1 and upper quartile Q 3 are: (2) Counting upper and lower abnormal points First, calculate the lower bound of the anomaly based on the upper and lower quartiles Q min With abnormal upper bound Q max : The lower abnormal point count is the amplitude of the original vibration signal not greater than Q min The number of points, the upper abnormal point count is the original vibration signal amplitude is not less than Q max The number of points.

[0026] (3) Normal amplitude histogram generation For the original vibration signal, the amplitude is Q min and Q max The data set between D ={ d 1, d 2, …, d m},in m ≤ n , set the number of groups b , generate a histogram of the magnitude h The horizontal axis of the histogram represents the range of each amplitude segment, and the vertical axis represents the frequency (number) of values ​​within this range. The amplitude range of the 300 points sampled each time is different. The histogram can eliminate the influence of the amplitude range and only generate the overall distribution information of the sampling data this time. Figure 3 The following is a normal amplitude histogram of a section of idle data ( b =18).

[0027] (1) Amplitude histogram stitching In the normal amplitude histogram, only the Q min and Q max The next step is to place the count value of the next outlier point between hAt the front end of h The end of b +2 groups of frequencies. Figure 4 Shown is a spliced ​​amplitude histogram of a segment of idle data.

[0028] (2) Histogram Normalization Divide the histogram counts obtained in step (4) by the total number of points to obtain a normalized amplitude histogram. Figure 5 shown.

[0029] 3. Histogram storage The sensors are fixed on various manufacturers of engineering machinery of different sizes and new and old and collect a large amount of idle data. Each sampling is performed through step 2 to generate the normalized histogram of this sampling. All normalized histograms are stored in the database as the prior distribution of idle data. w times, the size of the idle bank H is in b is the number of groups of the histogram described in step (3) of step 2, Figure 6 Shown is an example of partial data from a generated idle library.

[0030] 4. Operation status identification The operating state is identified by using a distance metric. For the newly collected vibration data, the corresponding normalized histogram is generated according to step 2. h new Next, calculate h new The Euclidean distance from all data in the idle library H: Final Generation w Euclidean distance ,calculate h new The average Euclidean distance to all data in the idle database Setting the distance threshold t , then the state 0 means idling and 1 means working.

[0031] Figure 7 The figure shows the recognition result of a certain segment of collected data. The upper part is the histogram similarity, and the lower part is the actual operating status. t =0.019, as shown by the red line. It can be seen from the figure that the threshold accurately divides the two operating states.

[0032] Example 2 This embodiment provides a device working time statistics system based on histogram distance measurement, including: The data acquisition module is configured to acquire a vibration signal of the mechanical equipment as idle speed data; A histogram module is configured to generate an amplitude histogram according to the acquired mechanical equipment vibration signal; The storage module is configured to store the generated amplitude histogram as a prior distribution of the idle data; The state recognition module is configured to use the stored amplitude histogram to perform state recognition of the vibration signal.

[0033] A computer-readable storage medium stores a plurality of instructions, wherein the instructions are suitable for being loaded and executed by a processor of a terminal device, for example, a method for counting equipment working hours based on histogram distance measurement.

[0034] A terminal device includes a processor and a computer-readable storage medium, wherein the processor is used to implement various instructions; the computer-readable storage medium is used to store multiple instructions, wherein the instructions are suitable for being loaded by the processor and executing the equipment working time statistics method based on histogram distance measurement.

[0035] The above are all preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for counting equipment working hours based on histogram distance measurement is characterized by: include: Obtain mechanical equipment vibration signals as idle data; Generate an amplitude histogram according to the acquired mechanical equipment vibration signal; The generated amplitude histogram is stored in the database as the prior distribution of the idle data; The stored amplitude histogram is used to identify the state of the vibration signal.

2. The equipment working time statistics method based on histogram distance measurement according to claim 1 is characterized in that: The obtaining of the mechanical equipment vibration signal includes collecting the original vibration signal of the mechanical equipment through a vibration signal sensor, with a collection frequency of 300 Hz, a sampling time of 1 s, and a sampling interval of 1 minute. The original vibration signal is sent to a cloud server through a gateway device using the MQTT protocol.

3. The equipment working time statistics method based on histogram distance measurement according to claim 2 is characterized in that: The method of generating an amplitude histogram based on the acquired mechanical equipment vibration signal includes obtaining quartiles by arranging the data of the vibration signal, counting upper and lower abnormal points according to the obtained quartiles, and generating a normal amplitude histogram for the data with amplitudes between the upper and lower abnormal points through counting the upper and lower abnormal points.

4. The equipment working time statistics method based on histogram distance measurement according to claim 3 is characterized in that: The method of generating an amplitude histogram based on the acquired mechanical equipment vibration signal also includes histogram splicing of normal amplitude histograms, wherein the lower abnormal point count value is placed at the front end of the histogram and the upper abnormal point count value is placed at the end of the histogram to obtain a spliced ​​amplitude histogram.

5. The equipment working time statistics method based on histogram distance measurement according to claim 4 is characterized in that: The step of generating an amplitude histogram based on the acquired mechanical equipment vibration signal also includes normalizing the spliced ​​amplitude histogram, wherein the histogram count of the spliced ​​amplitude histogram is divided by the total number of points to obtain a normalized amplitude histogram.

6. The equipment working time statistics method based on histogram distance measurement according to claim 5 is characterized in that: The generated amplitude histogram is stored in the library as the prior distribution of the idle data, including fixing sensors on engineering machinery of different sizes and new and old from various manufacturers and collecting a large amount of idle data, generating a normalized histogram of the sampling each time, storing all the normalized histograms in the library as the prior distribution of the idle data, assuming that a total of w times, the size of the idle bank H is in b is the number of groups in the histogram.

7. The equipment working time statistics method based on histogram distance measurement according to claim 6 is characterized in that: The state identification of the vibration signal is performed using the stored amplitude histogram, including calculating the Euclidean distance between the vibration signal and all the data in the idle library according to the amplitude histogram, thereby obtaining the average Euclidean distance between the vibration signal and all the data in the idle library, and obtaining the state identification result of the vibration signal by setting a threshold.

8. A equipment working time statistics system based on histogram distance measurement, characterized in that: include: The data acquisition module is configured to acquire a vibration signal of the mechanical equipment as idle speed data; A histogram module is configured to generate an amplitude histogram according to the acquired mechanical equipment vibration signal; The storage module is configured to store the generated amplitude histogram as a prior distribution of the idle data; The state recognition module is configured to use the stored amplitude histogram to perform state recognition of the vibration signal.

9. A computer-readable storage medium storing a plurality of instructions, characterized in that: The instructions are suitable for being loaded by a processor of a terminal device and executing the method according to claim 1 .

10. A terminal device, comprising a processor and a computer-readable storage medium, wherein the processor is used to implement each instruction; and the computer-readable storage medium is used to store multiple instructions, characterized in that: The instructions are suitable for being loaded by a processor and executing the method as claimed in claim 1 .

Images