A method and device for monitoring the operating status of a diesel generator set

Abstract

The present invention belongs to the technical field of state monitoring of diesel generator sets, and specifically relates to a method and device for monitoring the operating state of a diesel generator set, the method comprising: segmenting the historical sequence of each abnormal state of the diesel generator set, calculating the similarity between the sample sequence and each segment through the DTW algorithm and the median sequence of each segment, and obtaining the matching sequence number of the data corresponding to each real-time pressure in the sample sequence in the median sequence, obtaining the matching data with the same matching sequence number in the upper limit sequence and the lower limit sequence, correcting the similarity according to the size relationship between each real-time pressure in the sample sequence and the matching data, synthesizing the corrected similarity between sample sequences of different lengths and each segment, calculating the possibility that the operating state of the diesel generator set at the current moment corresponds to each segment, and taking the sequence number of the segment corresponding to the maximum possibility as the warning level at the current moment. The present invention realizes the monitoring of the operating state of the diesel generator set.

Description

Technical Field

[0001] The present invention relates to the technical field of state monitoring of diesel generator sets, and more specifically, to a method and device for monitoring the operating state of a diesel generator set. Background Art

[0002] Diesel generator sets are power generation equipment that rely on diesel as fuel to generate electricity. They are widely used in various fields, including industry, commerce, construction, hospitals and communications. The fuel system of a diesel generator set provides fuel supply for the start-up and operation of the diesel engine. Abnormal function of the fuel system will directly affect the use of the diesel generator set.

[0003] In the related technology, for example, a Chinese patent document with an authorization announcement number of CN116498435B discloses a method and device for monitoring the use status of a diesel generator set, including: calculating a distance vector between the operating data of the diesel generator set and a preset data matrix; calculating the vector weight of each vector element in the distance vector, and constructing the difference vector according to the vector weight; performing vector reconstruction on the difference vector to obtain a reconstructed vector, and generating a data feature map according to the reconstructed vector using a preset state monitoring model; allocating attention to the data feature map according to the attention layer in the state monitoring model to obtain an attention feature map of the operating data; performing a state warning on the diesel generator based on the attention feature map to obtain the real-time use status of the diesel generator, and recording events for the real-time use status. The real-time use status of the diesel generator set can be the use status of the diesel generator set such as electrical fault, mechanical fault, low-temperature water system fault, engine abnormality, and protective shutdown. The state of the diesel generator set is monitored in real time through the real-time use status, so that the staff can repair the state of the diesel generator set to ensure the normal operation of the diesel generator set.

[0004] The timeliness of identifying the operating status in the related technology is poor, and the operating status of the diesel generator set cannot be predicted in time, thereby failing to give the operator sufficient response time. Summary of the invention

[0005] In order to solve the technical problem that the timeliness of identifying the operating status in the above-mentioned related technologies is poor, the operating status of the diesel generator set cannot be predicted in time, and thus the operator cannot be given enough time to respond, the present invention provides solutions in the following aspects.

[0006] In a first aspect, the present invention provides a method for monitoring the operating status of a diesel generator set, comprising: obtaining the historical pressure of each measuring point of the diesel generator set; determining multiple moments corresponding to multiple state abnormalities in the historical pressure; intercepting multiple historical pressures before the moments to obtain multiple historical sequences, each state abnormality corresponding to a historical sequence; segmenting each historical sequence to obtain N segments, and the segments with the same sequence number have the same length; for the same segment of all historical sequences , segment The historical pressures with the same sequence number in the above example form a reference sequence. The number of all reference sequences is the same as the segment number. The same length; group the medians of all reference sequences into segments The median sequence of all reference sequences is grouped into segments according to the upper quartiles The upper limit sequence of all reference sequences is grouped into segments of the lower quartiles The lower limit sequence; collect the real-time pressure of each measuring point of the diesel generator set, intercept multiple real-time pressures before the current moment, and obtain the sample sequence; through the DTW algorithm and segmentation The median sequence of the sample sequence and the segmentation The similarity is obtained, and the matching sequence number of the data corresponding to each real-time pressure in the sample sequence in the median sequence is obtained; the matching data with the same matching sequence number in the upper limit sequence and the lower limit sequence is obtained; according to the size relationship between each real-time pressure in the sample sequence and the matching data, the similarity is corrected to obtain the sample sequence and the segmentation The corrected similarity of the sample sequences of different lengths and the corrected similarities of each segment is integrated to calculate the possibility that the operating status of the diesel generator set at the current moment corresponds to each segment, and the sequence number of the segment corresponding to the maximum possibility is used as the warning level at the current moment.

[0007] The present invention obtains the median sequence, upper limit sequence and lower limit sequence of each segment, providing a benchmark and reference for the operating status monitoring of the diesel generator set. These sequences help to establish the pressure range of each segment, ensure the reliability and accuracy of the prediction results, and thus can more effectively monitor and manage the operating status of the diesel generator set, respond to potential abnormal situations in a timely manner, and reduce the occurrence of failures and damages.

[0008] The present invention predicts the operating status of the diesel generator set at the current moment and divides the warning level, reflecting the distance between the current moment and the starting moment of the abnormal operating status of the diesel generator set, giving the operator enough response time, so that the operator can take necessary measures before the impact of the fault becomes serious, thereby realizing the monitoring of the operating status of the diesel generator set, thereby improving the operating efficiency and reducing the maintenance cost and safety hazards.

[0009] Preferably, the sample sequences of different lengths are determined according to the following method: when calculating the segment corresponding to the operating state of the diesel generator set at the current moment When the possibility of Each integer in is taken as a value of the length of the sample sequence, then the number of sample sequences of different lengths is equal to ; Indicates segmentation Length, Indicates rounding up.

[0010] By analyzing the pressure in different time periods, the present invention can better capture the changing trend of the real-time pressure of the diesel generator set during operation, and at the same time can help establish a comprehensive understanding of the state of the diesel generator set, so as to more accurately judge the current operating state.

[0011] Preferably, the calculation sample sequence and the segment The similarity includes: obtaining sample sequences and segments through the DTW algorithm The data correspondence relationship of the median sequence includes the data corresponding to each real-time pressure in the sample sequence in the median sequence; obtaining the data A1 with the largest sequence number among the multiple data corresponding to the first real-time pressure of the sample sequence, and obtaining the data A2 with the smallest sequence number among the multiple data corresponding to the last real-time pressure of the sample sequence; The sequence of all data between data A1 and data A2 in the median sequence is used as the sample sequence and segment The matching sequence of the median sequence includes data A1 and data A2; according to the ratio of the length of the matching sequence to the length of the sample sequence, the length of the sample sequence and the segment length are calculated. similarity.

[0012] Preferably, the sample sequence and the segment The similarity satisfies the expression: ; In the formula, Indicates the length is Sample sequence and segmentation The similarity of represents the length of the sample sequence, Indicates the length is Sample sequence and segmentation The length of the matching sequence of the median sequence, Indicates taking the absolute value, represents the natural exponential function.

[0013] Preferably, the similarity is corrected according to the size relationship between each real-time pressure in the sample sequence and the matching data to obtain the sample sequence and the segment The correction similarity includes: the matching data with the same matching sequence number as each real-time pressure in the upper limit sequence and the lower limit sequence refer to the upper limit data and the lower limit data corresponding to each real-time pressure, wherein the data with the same matching sequence number as each real-time pressure in the upper limit sequence is used as the upper limit data corresponding to each real-time pressure; the data with the same matching sequence number as each real-time pressure in the lower limit sequence is used as the lower limit data corresponding to each real-time pressure; according to the size relationship between each real-time pressure in the sample sequence and the matching data, the correction degree of the sample sequence is calculated; the product of the correction degree and the similarity is used as the similarity between the sample sequence and the segmented sample sequence. The modified similarity.

[0014] Preferably, the correction degree of the sample sequence satisfies the expression: , Indicates the length is The degree of correction of the sample sequence, represents the length of the sample sequence, Indicates taking the maximum value, Indicates the length is The first sample sequence of The abnormal degree of real-time pressure; and Satisfies the expression: , Indicates the length is The first sample sequence of Real-time pressure, , Respectively represent the length The first sample sequence of The mean of all upper limit data and the mean of all lower limit data corresponding to the real-time pressure, Indicates the length is The first sample sequence of The mean of all data corresponding to a real-time pressure in the median sequence.

[0015] The mean of all upper limit data corresponding to each real-time pressure in the present invention And the mean of all corresponding lower limit data , constituting the pressure range of the position where the real-time pressure is located. The abnormality degree of each real-time pressure is obtained through the size relationship between the real-time pressure and the pressure range of the position where the real-time pressure is located. Then, according to the abnormality degree of each real-time pressure in the sample sequence, the correction degree of the sample sequence is determined. The similarity between the sample sequence and each segment is corrected according to the correction degree, so as to improve the calculation accuracy of the similarity between the sample sequence and each segment.

[0016] Preferably, the probability that the operating state of the diesel generator set at the current moment corresponds to each segment satisfies the expression: ; In the formula, Indicates the current diesel generator set operating status corresponding to the segment The possibility of represents the length of the sample sequence, Indicates segmentation Length, Indicates rounding up. Indicates the length is Sample sequence and segmentation The modified similarity.

[0017] Preferably, the probability that the operating state of the diesel generator set at the current moment corresponds to each segment satisfies the expression: ; In the formula, Indicates the current diesel generator set operating status corresponding to the segment The possibility of represents the length of the sample sequence, Indicates segmentation Length, Indicates rounding up. Indicates the length is Sample sequence and segmentation The corrected similarity of Indicates the length is Sample sequence and segmentation The weighted coefficient of similarity.

[0018] The present invention predicts the early warning level of the operating status of the diesel generator set by calculating the probability of the operating status of the diesel generator set corresponding to each segment. Based on the predicted early warning level of the operating status of the diesel generator set, the maintenance and overhaul time can be dynamically adjusted, thereby reducing unnecessary maintenance and saving time and cost.

[0019] Preferably, the length is Sample sequence and segmentation The weighted coefficient of the similarity satisfies the expression: ; In the formula, Indicates the preset third quantity, represents the DTW distance, Indicates the length is Sample sequence and segmentation The matching sequence of the median sequence of Indicates the length is Sample sequence and segmentation The matching sequence of the median sequence of represents the natural exponential function.

[0020] The present invention obtains weighted coefficients of similarity between sample sequences of different lengths and each segment according to the distances between the matching sequences of the median sequence of each segment, thereby improving the calculation accuracy of the possibility that the operating state of the diesel generator set corresponds to each segment, thereby improving the prediction accuracy of the operating state of the diesel generator set.

[0021] In a second aspect, the present invention provides an operating status monitoring device for a diesel generator set, comprising a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the operating status monitoring method of the diesel generator set described above is implemented.

[0022] By adopting the above technical solution, the above-mentioned method for monitoring the operating status of a diesel generator set is generated into a computer program and stored in a memory so as to be loaded and executed by a processor, thereby making a terminal device based on the memory and the processor for easy use.

[0023] The beneficial effects of the present invention are:

[0024] The present invention obtains the median sequence, upper limit sequence and lower limit sequence of each segment, providing a benchmark and reference for the operating status monitoring of the diesel generator set. These sequences help to establish the pressure range of each segment, ensure the reliability and accuracy of the prediction results, and thus can more effectively monitor and manage the operating status of the diesel generator set, respond to potential abnormal situations in a timely manner, and reduce the occurrence of failures and damages.

[0025] Furthermore, the present invention predicts the operating status of the diesel generator set at the current moment and divides the warning level, reflecting the distance between the current moment and the starting moment of the abnormal operating status of the diesel generator set, giving the operator enough time to respond, so that the operator can take necessary measures before the impact of the fault becomes serious, thereby realizing the monitoring of the operating status of the diesel generator set, thereby improving operating efficiency and reducing maintenance costs and safety hazards. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:

[0027] Figure 1 is a flow chart schematically showing a method for monitoring the operating status of a diesel generator set in the present invention;

[0028] Figure 2It schematically shows the data correspondence between the sample sequence and the segmented median sequence, as well as the upper limit data and lower limit data corresponding to each real-time pressure in the sample sequence. DETAILED DESCRIPTION

[0029] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0030] The specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0031] The embodiment of the present invention discloses a method for monitoring the operating status of a diesel generator set, referring to Figure 1 , comprising steps S1 to S7:

[0032] S1. Collect the pressure of each measuring point in the fuel system pipeline of the diesel generator set.

[0033] The fuel system pipeline of the diesel generator set is divided into two parts: low-pressure fuel pipeline and high-pressure fuel pipeline according to the working pressure. The low-pressure fuel pipeline is provided with pressure head by the fuel booster pump, and the fuel in the fuel tank is transported to the inlet of the high-pressure fuel pump of the fuel system pipeline through the fuel filter and the overspeed protection slide valve. Part of the fuel is sucked in and pressurized by the high-pressure fuel pump, and sent to the injector through the high-pressure fuel pipeline. The injector atomizes the fuel and sprays it into the cylinder to complete the combustion work. The fuel not sucked in by the high-pressure fuel pump flows back to the fuel tank through the return pipeline and the return oil pipeline pressure regulating valve.

[0034] Specifically, measuring points are set at the inlet and outlet of the fuel boost pump, the inlet and outlet of the high-pressure oil pump, and the inlet and outlet of the return oil pipeline pressure regulating valve in the fuel system pipeline of the diesel generator set; the pressure of each measuring point is collected through the pressure transmitter connected to each measuring point; wherein the pressure transmitter is connected to the fuel system pipeline through the instrument pipeline, collects the pressure of each measuring point, and uploads it to the instrumentation and control cabinet of the diesel generator set, therefore, the pressure of each measuring point at each moment is stored in the instrumentation and control cabinet; the sampling frequency is 1 second / time.

[0035] S2. Obtain the historical pressure of the measuring point of the diesel generator set, obtain a historical sequence consisting of multiple historical pressures before each abnormal state, and segment each historical sequence to obtain multiple segments.

[0036] It should be noted that abnormal pressure in the fuel system pipeline of a diesel generator set will cause the diesel generator set to be unavailable. In order to maintain the reliable operation of the diesel generator set, it is necessary to identify the operating status before a diesel generator set failure occurs, so as to improve the operating efficiency of the diesel generator set and reduce maintenance costs and safety hazards. Therefore, the present invention selects the historical pressure of each measuring point of the diesel generator set before each abnormal state to form a reference sequence for predicting the operating status of the diesel generator set.

[0037] Specifically, for any measuring point of the diesel generator set, the historical pressure of the measuring point within the historical time period is obtained from the pressure of the measuring point at each moment stored in the instrumentation and control cabinet of the diesel generator set; in this embodiment, the historical time period refers to the three months after the diesel generator set is first used. In other embodiments, the historical time period can be set according to actual application scenarios and needs.

[0038] Furthermore, multiple moments corresponding to multiple state abnormalities are determined in all historical pressures of the measuring point, and the state abnormality refers to an abnormality in the operating state of the measuring point of the diesel generator set, such as excessive pressure, too low pressure, frequent pressure fluctuations, etc. Each state abnormality will last for a period of time. Therefore, each state abnormality includes a period of time between the start time and the end time, and the time difference between the start time of each state abnormality and the end time of the previous state abnormality is required to be greater than or equal to L; it should be noted that the state abnormality is determined and recorded by professional maintenance personnel.

[0039] Furthermore, L historical pressures before the moment corresponding to each state abnormality are intercepted to obtain multiple historical sequences, each state abnormality corresponds to a historical sequence, and the length of each historical sequence is equal to L; in this embodiment, the moment corresponding to each state abnormality refers to the starting moment of each state abnormality. In other embodiments, the moment corresponding to each state abnormality can be set according to actual application scenarios and requirements.

[0040] Among them, L is equal to a preset first number. The specific value of the preset first number can be set according to actual application scenarios and requirements. The value range of the preset first number is [120, 600]. The present invention sets the preset first number to 300.

[0041] Furthermore, each historical sequence is segmented by the Fisher optimal solution method to obtain N segments, and the segments with the same sequence number have the same length, and N is equal to a preset second number; wherein, the larger the sequence number of the segment, the closer the segment is to the starting time of the abnormal operating state of the diesel generator set.

[0042] It should be noted that the optimal way to segment the historical sequence can be obtained through the Fisher optimal solution method. The Fisher optimal solution method is a well-known technology and will not be described in detail here.

[0043] For example, the historical sequence corresponding to the first abnormal state is segmented, and the obtained N segments are segmented as follows: , Segment , ..., segmentation and segmentation , where segment , Segment , Segment , Segment They represent the first segment, the second segment, and the Segment, Segments; segment the historical sequence corresponding to the second state anomaly, and the obtained N segments are , Segment , ..., segmentation and segmentation , where segment , Segment , Segment and segmentation They represent the first segment, the second segment, and the Segment, Segments; Similarly, for the historical sequence segmentation corresponding to the Mth state anomaly, the N segments obtained are segments , Segment , ..., segmentation and segmentation , where segment , Segment , Segment and segmentation Respectively represent The first segment, the second segment, the Segment, segments; Indicates the total number of abnormalities in all states; , Segment , ..., segmentation The same length, segment , Segment , ..., segmentation The same length, segment , Segment , ..., segmentation The same length, segment , Segment , ..., segmentation The same length.

[0044] Among them, the specific value of the preset second number can be set according to the actual application scenario and requirements. It is required that the preset second number is smaller than the preset first number, and the value range of the preset second number is [3,7]. The present invention sets the preset second number to 5.

[0045] S3. Obtain the median sequence, upper limit sequence and lower limit sequence of each segment.

[0046] Specifically, for the same segment of all historical sequences , segment The historical pressure components of the same sequence number A reference sequence, the number of all reference sequences and the segment The amount of data is the same.

[0047] For example, the first segment in the segmentation result of the historical sequence corresponding to the first state abnormality is segment , segment for , , , , Respectively represent segments 1, 2, , Historical pressure, Indicates the length of the first segment; the first segment in the segmentation result of the historical sequence corresponding to the second state abnormality is the segment , segment for , , , , Respectively represent segments 1, 2, , historical pressure; similarly, The first segment in the segmentation result of the historical sequence corresponding to the secondary state abnormality is segment , segment for , , , , Respectively represent segments 1, 2, , All the historical pressures with the same sequence number in the first segment are combined into a reference sequence of the first segment, and a total of reference sequences, where the first reference sequence of the first segment obtained is , the second reference sequence of the first segment is obtained as , the first segment obtained The reference sequence is The reference sequence is , the last reference sequence of the first segment is obtained The reference sequence is .

[0048] Furthermore, for segmentation All reference sequences were analyzed, and the median, upper quartile and lower quartile of each reference sequence were obtained by box plot method; the medians of all reference sequences were grouped into segments The median sequence of all reference sequences is grouped into segments according to the upper quartiles The upper limit sequence of all reference sequences is grouped into segments of the lower quartiles The lower limit sequence of .

[0049] It should be noted that the box plot is a statistical chart used to show the dispersion of a set of data. The box plot method is a well-known technology and will not be described in detail here. By obtaining the median sequence, upper limit sequence and lower limit sequence of each segment, the pressure range of each segment can be established.

[0050] Exemplarily, the first reference sequence of the first segment is obtained by the box plot method the median of , upper quartile and lower quartile ; Obtain the second reference sequence of the first segment through the box plot method the median of , upper quartile and lower quartile ; Obtain the first segment through the box plot method Reference sequences the median of , upper quartile and lower quartile ; Obtain the first segment through the box plot method Reference sequences the median of , upper quartile and lower quartile ; The medians of all reference sequences are combined into the median sequence of the first segment, then the median sequence of the first segment is ; The upper quartiles of all reference sequences form the upper limit sequence of the first segment, then the upper limit sequence of the first segment is ; The lower quartiles of all reference sequences form the lower limit sequence of the first segment, then the lower limit sequence of the first segment is .

[0051] It should be noted that the present invention obtains the median sequence, upper limit sequence and lower limit sequence of each segment, providing a benchmark and reference for the operating status monitoring of the diesel generator set. These sequences help establish the pressure range of each segment, ensuring the reliability and accuracy of the prediction results, and thus more effectively monitoring and managing the operating status of the diesel generator set, responding to potential abnormal situations in a timely manner, and reducing the occurrence of failures and damages.

[0052] S4. Collect the real-time pressures of multiple measuring points of the diesel generator set, intercept multiple real-time pressures before the current moment, and obtain a sample sequence.

[0053] Specifically, in order to judge the operating status of the diesel generator set, multiple real-time pressures before the current moment are selected to form a sample sequence. The length of the real-time pressure component at the current moment is The sample sequence of represents the length of the sample sequence; where the length of the sample sequence Determine as follows: In the subsequent synthesis of sample sequences of different lengths and segments The modified similarity calculation of the diesel generator set operating state corresponding segment When the possibility of Each integer in is taken as a value of the length of the sample sequence, then the number of sample sequences of different lengths is equal to ; Indicates segmentation Length, Indicates rounding up.

[0054] It should be noted that during the operation of the diesel generator set, the pressure changes dynamically. By analyzing the pressure in different time periods, these changing trends can be better captured. At the same time, it can help establish a comprehensive understanding of the status of the diesel generator set, thereby more accurately judging the current operating status.

[0055] S5. Calculate the similarity between the sample sequence and each segment through the DTW algorithm and the median sequence of each segment.

[0056] Specifically, for any sample sequence at the current moment, the sample sequence and segmentation are obtained through the dynamic time warping algorithm (Dynamic Time Warping, DTW algorithm) The data correspondence relationship of the median sequence includes the data corresponding to each real-time pressure in the sample sequence in the median sequence, and the matching serial number of the data corresponding to each real-time pressure in the sample sequence in the median sequence is obtained. It should be noted that the data corresponding to each real-time pressure in the sample sequence in the median sequence can be one or more.

[0057] For example, the data correspondence between the sample sequence and the segmented median sequence is as follows: Figure 2 As shown, the data corresponding to the first real-time pressure in the sample sequence in the median sequence are the first data and the second data, so the matching numbers of the data corresponding to the first real-time pressure in the sample sequence in the median sequence are 1 and 2; the data corresponding to the second real-time pressure in the sample sequence in the median sequence are the third data and the fourth data, so the matching numbers of the data corresponding to the second real-time pressure in the sample sequence in the median sequence are 3 and 4; the data corresponding to the third real-time pressure in the sample sequence in the median sequence are the fifth data and the sixth data, so the matching numbers of the data corresponding to the third real-time pressure in the sample sequence in the median sequence are 5 and 6; the data corresponding to the fourth real-time pressure in the sample sequence in the median sequence is the seventh data, so the matching number of the data corresponding to the fourth real-time pressure in the sample sequence in the median sequence is 7.

[0058] It should be noted that the DTW algorithm is a method for calculating the similarity of two sequences by searching for the path with the minimum sum of matrix elements in the path from the first data to the last data in the two sequences. The total number of elements in the path is the DTW distance. The DTW algorithm allows for scaling on the time axis, that is, a certain data in one sequence can correspond to multiple data in the other sequence, and multiple data in one sequence can correspond to one data in the other sequence.

[0059] Further, the data A1 with the largest sequence number among the multiple data corresponding to the first real-time pressure of the sample sequence is obtained, and the data A2 with the smallest sequence number among the multiple data corresponding to the last real-time pressure of the sample sequence is obtained; The sequence of all data between data A1 and data A2 in the median sequence is used as the sample sequence and segment A matching sequence of the median sequence of , wherein the matching sequence includes data A1 and data A2;

[0060] Furthermore, according to the ratio of the length of the matching sequence to the length of the sample sequence, the length of the sample sequence and the segmentation are calculated. The similarity between the sample sequence and the segment The similarity satisfies the expression:

[0061] ;

[0062] In the formula, Indicates the length is Sample sequence and segmentation The similarity of represents the length of the sample sequence, Indicates the length is Sample sequence and segmentation The length of the matching sequence of the median sequence, Indicates taking the absolute value, represents the natural exponential function.

[0063] It should be noted that the longer the length of the matching sequence is, the more consistent it is with the length of the sample sequence, that is, the ratio of the two The closer it is to 1, the The smaller, The larger the length, the Sample sequence and segmentation The greater the similarity.

[0064] It should be noted that the median sequence of each segment represents the concentrated distribution trend of the pressure of each segment. Therefore, the similarity between the sample sequence and each segment is calculated based on the ratio of the length of the matching sequence to the length of the sample sequence, which can reflect whether the distribution trend of the real-time pressure in the sample sequence obeys the concentrated distribution trend of the pressure of each segment.

[0065] S6, obtain the matching data with the same matching sequence number in the upper limit sequence and the lower limit sequence; correct the similarity according to the size relationship between each real-time pressure in the sample sequence and the matching data, and obtain the sample sequence and segment The modified similarity.

[0066] Specifically, the matching data in the upper limit sequence and the lower limit sequence that are identical to the matching sequence number of each real-time pressure refer to the upper limit data and the lower limit data corresponding to each real-time pressure, wherein the data in the upper limit sequence that are identical to the matching sequence number of each real-time pressure are used as the upper limit data corresponding to each real-time pressure; the data in the lower limit sequence that are identical to the matching sequence number of each real-time pressure are used as the lower limit data corresponding to each real-time pressure; it should be noted that the upper limit data corresponding to each real-time pressure in the sample sequence can be one or more, and the corresponding lower limit data can be one or more.

[0067] For example, the upper limit data and the lower limit data corresponding to each real-time pressure are as follows: Figure 2As shown, since the matching sequence numbers of the data corresponding to the first real-time pressure in the sample sequence in the median sequence are 1 and 2, the upper limit data corresponding to the first real-time pressure in the sample sequence is the first data and the second data in the upper limit sequence; the data corresponding to the second real-time pressure in the sample sequence in the median sequence are the third data and the fourth data, so the matching sequence numbers of the data corresponding to the second real-time pressure in the sample sequence in the median sequence are 3 and 4; the data corresponding to the third real-time pressure in the sample sequence in the median sequence are the fifth data and the sixth data, so the matching sequence numbers of the data corresponding to the third real-time pressure in the sample sequence in the median sequence are 5 and 6; since the matching sequence number of the data corresponding to the fourth real-time pressure in the sample sequence in the median sequence is 7, the upper limit data corresponding to the fourth real-time pressure in the sample sequence is the seventh data in the upper limit sequence.

[0068] Further, according to the magnitude relationship between each real-time pressure in the sample sequence and the matching data, the correction degree of the sample sequence is calculated; the product of the correction degree and the similarity is used as the sample sequence and the segmentation The modified similarity.

[0069] The correction degree of the sample sequence satisfies the expression:

[0070] ;

[0071] In the formula, Indicates the length is The degree of correction of the sample sequence, represents the length of the sample sequence, Indicates taking the maximum value, Indicates the length is The first sample sequence of The abnormal degree of real-time pressure.

[0072] It should be noted that the length is The first sample sequence of Abnormal level of pressure is greater than 0, so The value range of is [0,1], then the length is The degree of correction of the sample sequence The value range is [0,1], and the degree of correction The smaller the sample sequence is, the closer the segment is to the The smaller the corrected similarity.

[0073] It should be further explained that the abnormal degree of each real-time pressure in the sample sequence is used to compare the sample sequence with the segmented The similarity of is modified, when the length is The first sample sequence of Abnormal degree of real-time pressure The larger the value, the more abnormal the real-time pressure is. At this time, the sample sequence and the segment length are calculated based on the ratio of the length of the matching sequence to the length of the sample sequence. The less credible the similarity, the greater the degree of abnormality. The larger the value, the more corrected the sample sequence is. The smaller the final sample sequence is, the smaller the The smaller the corrected similarity.

[0074] Among them, the length is The first sample sequence of Abnormal degree of real-time pressure Satisfies the expression:

[0075] ;

[0076] In the formula, Indicates the length is The first sample sequence of Real-time pressure, Indicates the length is The first sample sequence of The mean of all upper limit data corresponding to the pressure, Indicates the length is The first sample sequence of The mean of all lower limit data corresponding to the pressure, Indicates the length is The first sample sequence of The mean of all data corresponding to a real-time pressure in the median sequence.

[0077] It should be noted that for the length The first sample sequence of Real-time pressure , the mean of all upper limit data corresponding to the real-time pressure Represents the upper limit of the pressure range where the real-time pressure is located, and the mean of all lower limit data corresponding to the real-time pressure Represents the lower limit of the pressure range where the real-time pressure is located; when When the real-time pressure is within the pressure range of the position, it means that the real-time pressure is normal. Therefore, the abnormal degree of the real-time pressure ;when , the real-time pressure exceeds the upper limit of the pressure range at that location, indicating that the real-time pressure is abnormal, and the difference between the real-time pressure and the upper limit of the pressure range at that location The larger the value, the more abnormal the real-time pressure is. The bigger; when When the real-time pressure exceeds the lower limit of the pressure range at the location, it indicates that the real-time pressure is abnormal, and the difference between the real-time pressure and the lower limit of the pressure range at the location The larger the value, the more abnormal the real-time pressure is. The bigger.

[0078] It should be noted that the mean of all upper limit data corresponding to each real-time pressure in the present invention is And the mean of all corresponding lower limit data , constituting the pressure range of the position where the real-time pressure is located. The abnormality degree of each real-time pressure is obtained through the size relationship between the real-time pressure and the pressure range of the position where the real-time pressure is located. Then, according to the abnormality degree of each real-time pressure in the sample sequence, the correction degree of the sample sequence is determined. The similarity between the sample sequence and each segment is corrected according to the correction degree, so as to improve the calculation accuracy of the similarity between the sample sequence and each segment.

[0079] S7. Combining the sample sequences of different lengths with the corrected similarities of each segment, the probability of the operating state of the diesel generator set corresponding to each segment at the current moment is calculated, and the sequence number of the segment corresponding to the greatest probability is used as the warning level at the current moment.

[0080] It should be noted that the present invention uses sample sequences of different lengths to learn the pressure variation patterns on different time scales.

[0081] In one embodiment, the probability that the current diesel generator set operating state corresponds to each segment satisfies the expression:

[0082] ;

[0083] In the formula, Indicates the current diesel generator set operating status corresponding to the segment The possibility of represents the length of the sample sequence, Indicates segmentation Length, Indicates rounding up. Indicates the length is Sample sequence and segmentation The modified similarity.

[0084] It should be noted that all sample sequences of different lengths are The mean of the corrected similarity The larger the value is, the corresponding segment of the diesel generator set operation status at the current moment Possibility The bigger.

[0085] In another embodiment, the probability that the current diesel generator set operating state corresponds to each segment satisfies the expression:

[0086] ;

[0087] In the formula, Indicates the current diesel generator set operating status corresponding to the segment The possibility of represents the length of the sample sequence, Indicates segmentation Length, Indicates rounding up. Indicates the length is Sample sequence and segmentation The corrected similarity of Indicates the length is Sample sequence and segmentation The weighted coefficient of similarity.

[0088] Among them, the length is Sample sequence and segmentation The weighted coefficient of the similarity satisfies the expression:

[0089] ;

[0090] In the formula, Indicates the preset third quantity, represents the DTW distance, Indicates the length is Sample sequence and segmentation The matching sequence of the median sequence of Indicates the length is Sample sequence and segmentation The matching sequence of the median sequence of represents the natural exponential function.

[0091] Among them, the specific value of the preset third number can be set according to the actual application scenario and requirements, and the value range of the preset third number is [3,7]. The present invention sets the preset third number to 5.

[0092] It should be noted that the length is The sample sequence of length The sample sequence has increased real-time pressure, so if the length is The sample sequence determines the corresponding segment of the diesel generator set operation status , then except for the length Sample sequence and segmentation The larger the corrected similarity should be, the longer the length is The matching sequence corresponding to the sample sequence With length The matching sequence corresponding to the sample sequence Similarity should also be larger; among them, for the length The matching sequence corresponding to the sample sequence With length The matching sequence corresponding to the sample sequence , the DTW distance between the two The smaller the value, the more similar the two The bigger.

[0093] It should be noted that the present invention obtains the weighted coefficients of the similarities between sample sequences of different lengths and each segment based on the distances between the matching sequences of the median sequences of each segment, thereby improving the calculation accuracy of the possibility that the operating status of the diesel generator set corresponds to each segment, thereby improving the prediction accuracy of the operating status of the diesel generator set.

[0094] Furthermore, the corrected similarities of sample sequences of different lengths and each segment are integrated to calculate the probability that the operating status of the diesel generator set at the current moment corresponds to each segment, and the sequence number of the segment corresponding to the greatest probability is used as the warning level at the current moment, and the warning level is the prediction result; it should be noted that the sequence number of the warning level is the same as the sequence number of the segment. Since the larger the sequence number of the segment, the closer the segment is to the starting time of the abnormal operating status of the diesel generator set, therefore, the larger the sequence number of the warning level of the diesel generator set at the current moment, it means that the current moment is closer to the starting time of the abnormal operating status of the diesel generator set, which further indicates that the possibility of failure of the diesel generator set is increasing, the stability and reliability of the diesel generator set may decrease, and the risk increases.

[0095] Furthermore, the serial number of the warning level at the current moment can be regarded as a warning signal to prompt the operating personnel to strengthen the monitoring of the diesel generator set and take corresponding measures to prevent the occurrence of failures. Therefore, the present invention maintains and overhauls the diesel generator set based on the predicted warning level of the diesel generator set's operating status; compared to relying on a fixed maintenance and overhaul schedule, the present invention proposes dynamically adjusting the maintenance and overhaul time based on the prediction results, which can reduce unnecessary maintenance and save time and cost.

[0096] It should be noted that the present invention predicts the operating status of the diesel generator set at the current moment and divides the warning level, reflecting the distance between the current moment and the starting moment of the abnormal operating status of the diesel generator set, giving the operator enough time to respond, so that the operator can take necessary measures before the impact of the fault becomes serious, thereby realizing the monitoring of the operating status of the diesel generator set, thereby improving operating efficiency and reducing maintenance costs and safety hazards.

[0097] An embodiment of the present invention further discloses an operating status monitoring device for a diesel generator set, comprising a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, an operating status monitoring method for a diesel generator set according to the present invention is implemented.

[0098] The above system also includes other components well known to those skilled in the art, such as a communication bus and a communication interface, and their configuration and functions are known in the art, so they will not be described in detail here.

[0099] In the description of this specification, "plurality" or "several" means at least two, such as two, three or more, etc., unless otherwise clearly and specifically defined.

[0100] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, alternatives to the embodiments of the present invention described herein may be employed.

Claims

1. A method for monitoring the operating status of a diesel generator set, characterized in that: include: Obtain the historical pressure of each measuring point of the diesel generator set; determine multiple moments corresponding to multiple state abnormalities in the historical pressure; intercept multiple historical pressures before the moments to obtain multiple historical sequences, each state abnormality corresponds to a historical sequence; segment each historical sequence to obtain N segments, and the segments with the same sequence number have the same length; For the same segment of all historical sequences , segment The historical pressures with the same sequence number in the above example form a reference sequence. The number of all reference sequences is the same as the segment number. The same length; group the medians of all reference sequences into segments The median sequence of all reference sequences is grouped into segments according to the upper quartiles The upper limit sequence of all reference sequences is grouped into segments of the lower quartiles The lower limit sequence of ; Collect the real-time pressure of each measuring point of the diesel generator set, intercept multiple real-time pressures before the current moment, and obtain a sample sequence; through the DTW algorithm and segmentation The median sequence of the sample sequence and the segmentation and obtain the matching sequence number of the data corresponding to each real-time pressure in the sample sequence in the median sequence; obtain the matching data with the same matching sequence number in the upper limit sequence and the lower limit sequence; According to the size relationship between each real-time pressure in the sample sequence and the matching data, the similarity is corrected to obtain the sample sequence and segmentation The modified similarity of The corrected similarity of sample sequences of different lengths and each segment is combined to calculate the probability that the operating status of the diesel generator set at the current moment corresponds to each segment, and the sequence number of the segment corresponding to the maximum probability is used as the warning level at the current moment.

2. The method for monitoring the operating status of a diesel generator set according to claim 1, characterized in that: The sample sequences of different lengths are determined according to the following method: Calculate the corresponding segment of the diesel generator set operating status at the current moment When the possibility of Each integer in is taken as a value of the length of the sample sequence, then the number of sample sequences of different lengths is equal to ; Indicates segmentation Length, Indicates rounding up.

3. The method for monitoring the operating status of a diesel generator set according to claim 1, characterized in that: The calculation sample sequence and segment Similarities include: Obtain sample sequence and segmentation through DTW algorithm The data correspondence relationship of the median sequence includes the data corresponding to each real-time pressure in the sample sequence in the median sequence; obtaining the data A1 with the largest sequence number among the multiple data corresponding to the first real-time pressure of the sample sequence, and obtaining the data A2 with the smallest sequence number among the multiple data corresponding to the last real-time pressure of the sample sequence; The sequence of all data between data A1 and data A2 in the median sequence is used as the sample sequence and segment The matching sequence of the median sequence includes data A1 and data A2; according to the ratio of the length of the matching sequence to the length of the sample sequence, the length of the sample sequence and the segment length are calculated. similarity.

4. The method for monitoring the operating status of a diesel generator set according to claim 3, characterized in that: The sample sequence and segment The similarity satisfies the expression: ； In the formula, Indicates the length is Sample sequence and segmentation The similarity of represents the length of the sample sequence, Indicates the length is Sample sequence and segmentation The length of the matching sequence of the median sequence, Indicates taking the absolute value, represents the natural exponential function.

5. The method for monitoring the operating status of a diesel generator set according to claim 1, characterized in that: According to the size relationship between each real-time pressure in the sample sequence and the matching data, the similarity is corrected to obtain the sample sequence and the segment The modified similarity includes: The matching data in the upper limit sequence and the lower limit sequence that have the same matching sequence number as each real-time pressure refer to the upper limit data and the lower limit data corresponding to each real-time pressure, wherein the data in the upper limit sequence that have the same matching sequence number as each real-time pressure are used as the upper limit data corresponding to each real-time pressure; the data in the lower limit sequence that have the same matching sequence number as each real-time pressure are used as the lower limit data corresponding to each real-time pressure; the correction degree of the sample sequence is calculated according to the size relationship between each real-time pressure in the sample sequence and the matching data; the product of the correction degree and the similarity is used as the similarity between the sample sequence and the segmented sample sequence; The corrected similarity.

6. The method for monitoring the operating status of a diesel generator set according to claim 5, characterized in that: The correction degree of the sample sequence satisfies the expression: , Indicates the length is The degree of correction of the sample sequence, represents the length of the sample sequence, Indicates taking the maximum value, Indicates the length is The first sample sequence of The abnormal degree of real-time pressure; and Satisfies the expression: , Indicates the length is The first sample sequence of Real-time pressure, , Respectively represent the length The first sample sequence of The mean of all upper limit data and the mean of all lower limit data corresponding to the real-time pressure, Indicates the length is The first sample sequence of The mean of all data corresponding to a real-time pressure in the median sequence.

7. The method for monitoring the operating status of a diesel generator set according to claim 1, characterized in that: The probability that the current diesel generator set operating state corresponds to each segment satisfies the expression: ； In the formula, Indicates the current diesel generator set operating status corresponding to the segment The possibility of represents the length of the sample sequence, Indicates segmentation Length, Indicates rounding up. Indicates the length is Sample sequence and segmentation The corrected similarity.

8. The method for monitoring the operating status of a diesel generator set according to claim 1, characterized in that: The probability that the current diesel generator set operating state corresponds to each segment satisfies the expression: ； In the formula, Indicates the current diesel generator set operating status corresponding to the segment The possibility of represents the length of the sample sequence, Indicates segmentation Length, Indicates rounding up. Indicates the length is Sample sequence and segmentation The corrected similarity of Indicates the length is Sample sequence and segmentation The weighted coefficient of similarity.

9. The method for monitoring the operating status of a diesel generator set according to claim 3 or 8, characterized in that: The length is Sample sequence and segmentation The weighted coefficient of the similarity satisfies the expression: ； In the formula, Indicates the preset third quantity, represents the DTW distance, Indicates the length is Sample sequence and segmentation The matching sequence of the median sequence of Indicates the length is Sample sequence and segmentation The matching sequence of the median sequence of represents the natural exponential function.

10. A device for monitoring the operating status of a diesel generator set, characterized in that: include: A processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a method for monitoring the operating status of a diesel generator set according to any one of claims 1 to 9 is implemented.

Images