Comprehensive management and optimization system of industrial internet

Abstract

The invention discloses a comprehensive management and optimization system of an industrial internet, and relates to the technical field of industrial equipment management.The historical fault data of a plurality of continuous historical periods are extracted from a maintenance transmission log and are analyzed and processed to obtain a historical high fault set; performing high fault frequency value analysis processing on a plurality of fault high-frequency periods in a historical high fault set to obtain a sub-period comparison value, so as to reflect the overall characteristics of the same-rank historical high-frequency sub-periods after different fault high-frequency periods are divided in two aspects of fault frequency variation stability and trend similarity through the sub-period comparison value; the fault rule of the historical high-frequency sub-time periods with the same ranking after different fault high-frequency time periods are divided can be identified, and data support is provided for maintenance of industrial equipment in a future period.

Description

Technical Field

[0001] The present invention relates to the technical field of industrial equipment management, and particularly relates to an integrated management and optimization system for industrial Internet. Background Art

[0002] In the field of industrial production, the stable operation of industrial equipment is crucial for production efficiency, product quality, and the economic benefits of enterprises. With the rapid development of industrial Internet technology, it has become an inevitable trend to realize the intelligent management of the production process of industrial equipment.

[0003] In the prior art, the operation faults of industrial equipment are usually dealt with only when faults occur, lacking the evaluation of faults based on the behavior data of industrial equipment. Therefore, by analyzing the stability of the change trends of multiple historical fault data, this application is beneficial for enterprises (maintenance personnel) to plan and manage the predicted fault frequency of future equipment, and helps to identify the characteristic laws of the changes of multiple historical fault data. After determining its stability, based on this, predict the fault probability of industrial equipment in future cycle periods, and complete the prediction work of the fault frequency in future periods, thereby being beneficial for enterprises to efficiently manage industrial equipment.

[0004] Therefore, the present invention provides an integrated management and optimization system for industrial Internet. Summary of the Invention

[0005] The purpose of the present invention is to provide an integrated management and optimization system for industrial Internet to solve at least one of the above-mentioned prior art problems.

[0006] An integrated management and optimization system for industrial Internet includes the following steps: High-demand screening module: Analyze and process the historical fault data and corresponding historical periods within multiple consecutive historical cycles to obtain a historical high-fault set; Among them, divide the historical cycle into several consecutive historical time periods with equal time lengths, and record the historical fault frequency greater than or equal to the rated fault frequency within the historical time period as a high-fault frequency value; Mark the historical time period corresponding to the high-fault frequency value as a high-fault frequency time period; Type analysis module: Construct an analysis matrix for the high-fault frequency time periods and high-fault frequency values within the historical high-fault set, process the historical fault frequency data within the analysis matrix, and output the change types of sub-time periods; Regularity analysis module: Process the regularities of the stable change sub-time periods and fluctuating change sub-time periods that appear within multiple high-fault frequency time periods to obtain a periodic change signal or a non-periodic change signal based on the faults of industrial equipment; Prediction processing module: Based on periodic change signals and aperiodic change signals, obtain the predicted high-frequency effective values under different periodic changes to complete the prediction of industrial equipment failures in future time periods.

[0007] A further solution of the present invention: The method for obtaining the historical high-fault set is as follows: Sort multiple historical cycles according to the time sequence of the maintenance delivery log and form a historical set T; Based on any historical cycle in the historical set T, integrate and statistically analyze multiple fault high-frequency time periods according to the time sequence of the historical cycle to obtain the historical high-fault set.

[0008] A further solution of the present invention: The method for obtaining the analysis matrix is as follows: In the historical high-fault set, divide each fault high-frequency time period into several historical high-frequency sub-time periods with equal time lengths; Perform matrix sorting processing on the historical high-frequency sub-time periods divided for each fault high-frequency time period to obtain the analysis matrix , where the row data in the analysis matrix R represents the historical high-frequency sub-time periods divided within each fault high-frequency time period, the column data in the analysis matrix R represents the historical high-frequency sub-time periods divided within different fault high-frequency time periods, n represents the total number of historical high-frequency sub-time periods within each fault high-frequency time period, and m represents the total number of fault high-frequency time periods.

[0009] A further solution of the present invention: The quantization process of the historical fault frequency data in the analysis matrix is as follows: Extract the column data in the analysis matrix R to obtain the set ; Obtain The historical fault frequencies corresponding to the two endpoints of the historical high-frequency sub-time period, sum them up and take the average to obtain the unit change average value; Traverse all the historical high-frequency sub-time periods in the set in accordance with the processing method of the historical high-frequency sub-time period until the acquisition of the unit change average values corresponding to all the historical high-frequency sub-time periods in the set is completed; Calculate the variance of all the unit change average values to obtain the change stability value.

[0010] A further solution of the present invention: The quantization process of the historical fault frequency data in the analysis matrix is as follows: Arbitrarily combine adjacent historical high-frequency sub-time periods in the set and sort them according to the time sequence to obtain the element analysis group; Extract the historical high-frequency sub-period with the earliest time within the element analysis group, and obtain the historical failure frequencies corresponding to the two endpoints of the historical high-frequency sub-period with the earliest time. Substitute the historical failure frequencies corresponding to the two endpoints of the historical high-frequency sub-period with the earliest time into the two-dimensional coordinate system to obtain the change line for the earliest period; Based on the change line for the earliest period, calculate the corresponding slope through the slope formula ; Similarly, extract the historical high-frequency sub-period with the latest time within the element analysis group, and calculate the corresponding slope through the slope formula ; Subtract the slope from the slope , take the absolute value, and obtain the combined slope difference; Sum up the combined slope differences corresponding to all element analysis groups and take the average value to obtain the trend similarity value.

[0011] A further solution of the present invention: The process of outputting the change type of the sub-period is as follows: Sum up the trend similarity value and the change stability value to obtain the sub-period comparison value; If the sub-period comparison value is greater than the sub-period comparison threshold, generate a period fluctuation signal, and mark the historical high-frequency sub-period corresponding to the generated period fluctuation signal as the change fluctuation sub-period; If the sub-period comparison value is less than or equal to the sub-period comparison threshold, generate a period stability signal, and mark the historical high-frequency sub-period corresponding to the generated period stability signal as the change stability sub-period.

[0012] A further solution of the present invention: The method for obtaining the rule analysis data is as follows: Arbitrarily select a failure high-frequency period; Sort all the change stability sub-periods and change fluctuation sub-periods within the failure high-frequency period in chronological order to obtain a sub-period sequence; Obtain the number of change fluctuation sub-periods between adjacent change stability sub-periods, sum them up and take the average value, and calculate the ratio with the total number of high-demand sub-periods within the failure high-frequency period to obtain the change interval number; Combine adjacent failure high-frequency periods into a group of rule analysis groups, and sort them in chronological order of adjacent failure high-frequency periods within the rule analysis group to obtain multiple groups of rule analysis groups; Based on multiple groups of rule analysis groups, calculate the failure frequency interval characterization value through the Euclidean distance .

[0013] A further solution of the present invention: The method for obtaining the rule analysis data is as follows: Within the sub-period sequence, take the high fault frequency values corresponding to adjacent stable change sub-periods, subtract them, take the absolute value, and obtain the adjacent unit change value; Extract the maximum adjacent unit change value and the minimum adjacent unit change value, subtract them, and obtain the adjacent unit change value range; Add the maximum adjacent unit change value and the minimum adjacent unit change value, calculate the average value, and calculate the ratio with the adjacent unit change value range to obtain the stable change value; Calculate the high fault frequency characterization value by calculating the Euclidean distance for the stable change value 。

[0014] A further solution of the present invention: The process of identifying periodic changes is as follows: The fault frequency interval characterization value and the high fault frequency characterization value are added together to obtain the rule analysis value; If the rule analysis value is greater than the rule analysis threshold, a non-periodic change signal is generated; If the rule analysis value is less than or equal to the rule analysis threshold, a periodic change signal is generated.

[0015] A further solution of the present invention: The method for obtaining the predicted high-frequency effective value is as follows: Based on the periodic change signal, if the future period is a stable change period, obtain the average value of the unit changes corresponding to the historical high-frequency sub-periods, add them together and take the average value to obtain the predicted high-frequency effective value; If the future period is a fluctuating change period, perform prediction by the moving average method to obtain the predicted high-frequency effective value corresponding to the future period; Based on the non-periodic change signal, extract the maximum high fault frequency value and the minimum high fault frequency value in the analysis matrix R, add them together and take the average value to obtain the predicted high-frequency effective value corresponding to the future period.

[0016] The beneficial effects of the present invention: 1. The present invention extracts historical fault data of multiple consecutive historical periods from the maintenance delivery log, analyzes and processes it to obtain the historical high fault set, analyzes the high fault frequency values of multiple fault high-frequency periods in the historical high fault set to obtain the historical high-demand analysis data, and performs quantization processing to obtain the sub-period comparison value. Thus, through the sub-period comparison value, the overall characteristics of the historical high-frequency sub-periods ranked equally after being divided by different fault high-frequency periods in terms of the stability of the fault frequency change amount and the similarity of the trends are reflected, which is beneficial to identifying the fault rules of the historical high-frequency sub-periods ranked equally after being divided by different fault high-frequency periods, and provides data support for the maintenance of industrial equipment in the future period; 2. From multiple high-frequency fault time periods, any one is selected, and the internal stable change sub-periods and fluctuating change sub-periods are sorted in chronological order to form a sub-period sequence. The number of fluctuating change sub-periods between adjacent stable change sub-periods is counted to obtain a fault frequency interval characterization value. Also, the absolute value of the difference between the high fault frequency values corresponding to adjacent stable change sub-periods is taken to obtain an adjacent unit change value. Then, both are calculated through the Euclidean formula to obtain a fault frequency interval characterization value and a high fault frequency characterization value. The fault frequency interval characterization value and the high fault frequency characterization value are added together for calculation to obtain a pattern analysis value, thus realizing the quantitative analysis of the appearance pattern of sub-periods and the stability of fault frequency changes within multiple high-frequency fault time periods, identifying the change pattern of historical industrial equipment faults in the time series, and based on the obtained change pattern, it helps equipment maintenance personnel reasonably plan the maintenance plan; 3. By predicting the fault frequency for future use for the generated periodic change signal and aperiodic change signal, a predicted high-frequency effective value is obtained to complete the prediction work of the fault frequency for future time periods. On the one hand, when generating a periodic change signal, through the method of classification discussion, the prediction process is respectively carried out for the future time period being a stable change period and the future time period being a fluctuating change period to obtain a predicted high-frequency effective value. On the other hand, when generating an aperiodic change signal, the maximum high fault frequency value and the minimum high fault frequency value in the analysis matrix R are obtained, and their sum is taken and averaged to obtain the predicted high-frequency effective value corresponding to the future time period, which is beneficial for equipment maintenance personnel to reasonably arrange the maintenance time, accurately plan and manage the operation status of future industrial equipment, and helps to avoid the situation where equipment faults cannot be maintained in time. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0018] Figure 1 It is a flowchart of a comprehensive management and optimization method for an industrial Internet of the present invention; Figure 2 It is a schematic diagram of a comprehensive management and optimization system for an industrial Internet of the present invention; Figure 3 It is a schematic diagram of the structure of a comprehensive management and optimization device for an industrial Internet of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0019] To enable those skilled in the art to better understand the solution of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative work shall fall within the protection scope of the present invention.

[0020] Embodiment 1 Figure 1 As shown in the flowchart of a comprehensive management and optimization method for industrial Internet provided in Embodiment 1 of the present invention, this embodiment of the present invention is applicable to the screening, extraction, integration, and induction of high failure frequency values of industrial equipment based on multiple historical data to obtain a historical high failure set, and to analyze the stability of the characteristic laws of high failure frequency values based on the historical high failure set. This comprehensive management and optimization method for industrial Internet can be executed by a comprehensive management and optimization system for industrial Internet, and this comprehensive management and optimization system for industrial Internet can be implemented by software and / or hardware. This comprehensive management and optimization method for industrial Internet can be configured in a comprehensive management and optimization device for industrial Internet. Optionally, a comprehensive management and optimization device for industrial Internet can be an electronic device, and this electronic device can be a notebook, a desktop computer, a smart tablet, etc. The embodiments of the present invention do not limit this.

[0021] As Figure 1 shown, a comprehensive management and optimization method for industrial Internet provided in an embodiment of the present invention specifically includes the following steps: Step 1: Obtain historical failure data of industrial equipment in multiple consecutive historical periods through maintenance delivery logs, and analyze and process the historical failure data and the corresponding historical periods to obtain a historical high failure set; It should be noted that the historical periods include but are not limited to 10 days, 30 days, or 60 days, and the historical failure data includes mechanical failures, electrical failures, and control and software failures of industrial equipment; In some embodiments, multiple historical periods are sorted according to the time sequence of the maintenance delivery logs and form a historical set T = { , , ,...... }, where represents the nth historical period, and n represents the total number of historical periods; Based on any historical period in the historical set T, divide the historical period into several consecutive historical time segments with equal time lengths, obtain the historical fault frequency within the historical time segments, compare the magnitudes of the historical fault frequencies within all historical time segments, and record the values of the historical fault frequencies greater than or equal to the rated fault frequency as high fault frequency values; The rated fault frequency of the historical time segment is set by the staff based on experience; Mark the historical time segment corresponding to the high fault frequency value as the high fault frequency time segment; It should be noted that the historical fault frequency within the historical time segment represents the ratio of the total number of faults within the historical time segment to the duration of the historical time segment; Integrate and statistically analyze multiple high fault frequency time segments in the chronological order of the historical period to obtain the historical high fault set; Step 2: Based on the historical high fault set, analyze the high fault frequency values of multiple high fault frequency time segments to obtain an analysis matrix, perform quantization processing on the historical fault frequency data within the analysis matrix to obtain sub - time segment comparison values, compare them with the sub - time segment comparison threshold, and output to obtain the stable change sub - time segment and the fluctuating change sub - time segment; In some embodiments, in the historical high fault set, each high fault frequency time segment is divided into several historical high - frequency sub - time segments with equal time lengths, where the division method for each high fault frequency time segment is the same; Perform matrix sorting processing on the historical high - frequency sub - time segments divided for each high fault frequency time segment to obtain the analysis matrix , where the row data in the analysis matrix R represents the historical high - frequency sub - time segments divided within each high fault frequency time segment (one row represents one high fault frequency time segment), the column data in the analysis matrix R represents the historical high - frequency sub - time segments divided within different high fault frequency time segments (one column represents the historical high - frequency sub - time segments with the same ranking within different high fault frequency time segments), n represents the total number of historical high - frequency sub - time segments within each high fault frequency time segment, and m represents the total number of high fault frequency time segments; Exemplarily, extract the column data in the analysis matrix R to obtain the set ; Obtain the historical fault frequencies corresponding to the two endpoints of the historical high - frequency sub - time segment, sum them up and take the average to obtain the unit change average value; Traverse all the historical high - frequency sub - time segments within the set in accordance with the processing method of the historical high - frequency sub - time segment until the acquisition of the unit change average values corresponding to all the historical high - frequency sub - time segments within the set is completed; Calculate the variance of all the unit change average values to obtain the change stability value; Arbitrarily within the set Combine adjacent historical high-frequency sub-periods within it and sort them according to the time series to obtain an element analysis group; Extract the historical high-frequency sub-period with an earlier time within the element analysis group, and obtain the historical fault frequencies corresponding to the two endpoints of the historical high-frequency sub-period with an earlier time. Establish a two-dimensional coordinate system, where the X-axis is time and the Y-axis is the fault frequency. Substitute the historical fault frequencies corresponding to the two endpoints of the historical high-frequency sub-period with an earlier time into the two-dimensional coordinate system to obtain a change line for the earlier period; Similarly, extract the historical high-frequency sub-period with a later time within the element analysis group, and obtain the historical fault frequencies corresponding to the two endpoints of the historical high-frequency sub-period with a later time. Establish a two-dimensional coordinate system, where the X-axis is time and the Y-axis is the fault frequency. Substitute the historical fault frequencies corresponding to the two endpoints of the historical high-frequency sub-period with a later time into the two-dimensional coordinate system to obtain a change line for the later period; Based on the change line for the earlier period, through the slope calculation formula: Calculate the slope corresponding to the change line for the earlier period , where, ( , ) and ( , ) represent the coordinates corresponding to the two endpoints of the historical high-frequency sub-period with an earlier time; Based on the change line for the later period, through the slope calculation formula: Calculate the slope corresponding to the change line for the later period , where, ( , ) and ( , ) represent the coordinates corresponding to the two endpoints of the historical high-frequency sub-period with a later time; Take the difference between the slope and the slope , take the absolute value to obtain a combined slope difference; Sum and average the combined slope differences corresponding to all element analysis groups to obtain a trend similarity value; Multiply the trend similarity value by the change stability value to obtain a sub-period comparison value; It can be explained that the meaning represented by the sub-period comparison value is: comprehensively reflecting the overall characteristics of the historical high-frequency sub-periods of the same ranking after being divided by different fault high-frequency periods in terms of the stability of the fault frequency change amount and the similarity of the trend, which is conducive to identifying the fault laws of the historical high-frequency sub-periods of the same ranking after being divided by different fault high-frequency periods, and providing data support for equipment maintenance in the future cycle; Compare the sub-period comparison value with the sub-period comparison threshold, and the process is as follows: If the comparison value of the sub-period is greater than the comparison threshold of the sub-period, it indicates that the change in the average failure frequency between adjacent sub-periods is relatively large, and a period fluctuation signal is generated. The historical high-frequency sub-period corresponding to the generated period fluctuation signal is marked as a changing and fluctuating sub-period; If the comparison value of the sub-period is less than or equal to the comparison threshold of the sub-period, it indicates that the change in the average failure frequency between adjacent sub-periods is relatively small, and a period stability signal is generated. The historical high-frequency sub-period corresponding to the generated period stability signal is marked as a changing and stable sub-period; Exemplarily, analyzing whether the historical high-frequency sub-period is a changing and fluctuating sub-period or a changing and stable sub-period aims to, for the historical high-frequency sub-periods with the same time series within each high-failure period, the stability of the corresponding failure frequency change amount and the similarity of trends are beneficial to identifying the failure laws of the historical high-frequency sub-periods ranked the same after being divided by different high-failure periods, providing data support for equipment maintenance in future cycles; The specific implementation manner of the embodiment of the present invention is as follows: By extracting historical failure data of multiple consecutive historical cycles from the maintenance delivery log, analyzing and processing it to obtain a historical high-failure set, analyzing the high-failure frequency values of multiple high-failure periods within the historical high-failure set to obtain historical high-data to be analyzed, and performing quantization processing to obtain a sub-period comparison value, so as to reflect the overall characteristics of the historical high-frequency sub-periods ranked the same after being divided by different high-failure periods in terms of the stability of the failure frequency change amount and the similarity of trends through the sub-period comparison value, which is beneficial to identifying the failure laws of the historical high-frequency sub-periods ranked the same after being divided by different high-failure periods, providing data support for equipment maintenance in future cycles.

[0022] Embodiment 2 Such as Figure 1 As shown, on the basis of Embodiment 1, a comprehensive energy management and optimization method based on artificial intelligence provided by the embodiment of the present invention specifically includes: Step 3: Analyze the occurrence laws of the changing and stable sub-periods and the changing and fluctuating sub-periods within multiple high-failure periods to obtain law analysis data. Among them, the law analysis data includes a failure frequency interval characterization value and a high-failure frequency characterization value. Perform quantization processing on the law analysis data to obtain a law analysis value, and compare it with a law analysis threshold to obtain a law analysis result; Among them, the law analysis result includes a periodic change signal or a non-periodic change signal; In some embodiments, arbitrarily select a high-failure period; Sort all the changing and stable sub-periods and the changing and fluctuating sub-periods within the high-failure period in chronological order to obtain a sub-period sequence; Within the sub - time - period sequence, obtain the number of change - fluctuation sub - time - periods between adjacent stable - change sub - time - periods, sum them up and take the average, and calculate the ratio with the total number of high - demand sub - time - periods within the high - frequency fault time - periods to obtain the change - interval number. Combine adjacent high - frequency fault time - periods into a group of pattern - analysis groups, and sort them according to the time sequence of adjacent high - frequency fault time - periods within the pattern - analysis groups to obtain multiple groups of pattern - analysis groups. Substitute the change - interval numbers corresponding to adjacent high - frequency fault time - periods into the Euclidean - distance calculation formula: to calculate the fault - frequency - interval characterization value , where n represents the total number of pattern - analysis groups, i represents the sorting of the pattern - analysis group within the multiple groups of pattern - analysis groups, C i represents the change - interval number of the i - th pattern - analysis group, and C i-1 represents the change - interval number of the (i - 1) - th pattern - analysis group. Within the sub - time - period sequence, take the high - fault - frequency values corresponding to adjacent stable - change sub - time - periods, subtract them, take the absolute value to obtain the adjacent - unit change value. Compare the magnitudes of all adjacent - unit change values, extract the maximum adjacent - unit change value and the minimum adjacent - unit change value, and subtract them to obtain the adjacent - unit change - value range. Sum up and take the average of the maximum adjacent - unit change value and the minimum adjacent - unit change value, and calculate the ratio with the adjacent - unit change - value range to obtain the stable - change value. Substitute the stable - change values corresponding to adjacent high - frequency fault time - periods into the Euclidean - distance calculation formula: to calculate the high - fault - frequency characterization value , where n represents the total number of pattern - analysis groups, i represents the sorting of the pattern - analysis group within the multiple groups of pattern - analysis groups, E i represents the stable - change value of the i - th pattern - analysis group, and E i-1 represents the stable - change value of the (i - 1) - th pattern - analysis group. Add the fault - frequency - interval characterization value and the high - fault - frequency characterization value to obtain the pattern - analysis value. It can be understood that the meaning represented by the rule analysis value is as follows: it reflects the stability of the intervals between the stable sub-periods and the fluctuating sub-periods during multiple high-fault-frequency periods, as well as the stability of the changes in the high-fault-frequency values between adjacent stable sub-periods. Specifically, it reflects the relative stability degree of the number of fluctuating sub-periods between adjacent stable sub-periods through the fault-frequency interval characterization value, and reflects the relative stability degree of the change amplitude of the high-fault-frequency values between adjacent stable sub-periods through the high-fault-frequency characterization value. The rule analysis value obtained by summing the two comprehensively characterizes the stability of the relevant periods and the changes in fault frequency during multiple high-fault-frequency periods; Compare the rule analysis value with the rule analysis threshold, and the process is as follows: If the rule analysis value is greater than the rule analysis threshold, it indicates that the fluctuation degree of the intervals between the stable sub-periods and the fluctuating sub-periods during multiple high-fault-frequency periods is relatively large, and the fluctuation degree of the changes in the high-fault-frequency values between adjacent stable sub-periods is relatively large, and an aperiodic change signal is generated; If the rule analysis value is less than or equal to the rule analysis threshold, it indicates that the fluctuation degree of the intervals between the stable sub-periods and the fluctuating sub-periods during multiple high-fault-frequency periods is relatively small, and the fluctuation degree of the changes in the high-fault-frequency values between adjacent stable sub-periods is relatively small, and a periodic change signal is generated; The specific implementation scheme of the embodiment of the present invention is as follows: Select any one from multiple high-fault-frequency periods, sort the stable sub-periods and fluctuating sub-periods inside it in chronological order to form a sub-period sequence, count the number of fluctuating sub-periods between adjacent stable sub-periods to obtain the fault-frequency interval characterization value, and take the absolute value of the difference between the high-fault-frequency values corresponding to adjacent stable sub-periods to obtain the adjacent unit change value. Then calculate the two respectively through the Euclidean formula to obtain the fault-frequency interval characterization value and the high-fault-frequency characterization value, and add the fault-frequency interval characterization value and the high-fault-frequency characterization value to calculate the rule analysis value, thereby realizing the quantitative analysis of the appearance rules of sub-periods and the stability of fault-frequency changes during multiple high-fault-frequency periods, identifying the change patterns of historical equipment fault maintenance requirements in the time series, and based on the obtained change patterns, it helps maintenance personnel to reasonably maintain the equipment.

[0023] Embodiment III As Figure 1 shown, on the basis of Embodiment 1 and Embodiment 2, the embodiment of the present invention provides a comprehensive energy management and optimization method based on artificial intelligence; Step Four: Based on the periodic change signal and the aperiodic change signal, obtain the predicted high-frequency effective value under different periodic changes to complete the prediction of industrial equipment faults in future periods; In some embodiments, based on the periodic change signal, if the future time period is a stable change period, the average unit change corresponding to the historical high-frequency sub-period is obtained, summed up and averaged to obtain the predicted high-frequency effective value; If the future time period is a fluctuating change period, prediction is performed by the moving average method, and the process is as follows: S1. According to the time sequence corresponding to the time period, extract the high fault frequency values corresponding to the first three historical high-frequency sub-periods before the future time period; S2. Sum up and average the high fault frequency values corresponding to the first three historical high-frequency sub-periods before the future time period to obtain the predicted high-frequency effective value corresponding to the future time period; Based on the aperiodic change signal, the process of obtaining the predicted high-frequency effective value is as follows: Based on the analysis matrix R ; Extract the maximum high fault frequency value and the minimum high fault frequency value in the analysis matrix R, sum them up and average to obtain the predicted high-frequency effective value corresponding to the future time period; It should be noted that the significance of obtaining the predicted high-frequency effective value is as follows: it helps industrial equipment to understand the equipment fault conditions in different future time periods in advance. For industrial equipment maintenance, it means reasonably arranging maintenance resources according to the predicted high-frequency effective value to ensure that there are sufficient maintenance personnel and maintenance equipment during the equipment fault period; The specific implementation manner of the embodiment of the present invention is as follows: by generating periodic change signals and aperiodic change signals, predicting the future fault frequency, obtaining the predicted high-frequency effective value, and completing the prediction of the future time period fault frequency. On the one hand, when generating periodic change signals, through the method of classified discussion, prediction processing is respectively performed on the future time period being a stable change period and the future time period being a fluctuating change period to obtain the predicted high-frequency effective value. On the other hand, when generating aperiodic change signals, the maximum high fault frequency value and the minimum high fault frequency value in the analysis matrix R are obtained, summed up and averaged to obtain the predicted high-frequency effective value corresponding to the future time period, which is beneficial for equipment maintenance personnel to reasonably arrange maintenance time, accurately plan and manage the operation status of future industrial equipment, and help avoid the situation that equipment faults cannot be maintained in time.

[0024] Embodiment 4 As Figure 2 shown, on the basis of Embodiment 1, Embodiment 2 and Embodiment 3, an integrated management and optimization system for industrial Internet provided by the embodiment of the present invention includes: High-demand screening module: Obtain historical fault data within multiple consecutive historical cycles through the maintenance delivery log, analyze and process the historical fault data to obtain a historical set, analyze the historical cycles in the historical set to obtain a historical high-fault set; Type analysis module: Based on the historical high-fault set, analyze the high-fault frequency values for multiple high-fault time periods to obtain an analysis matrix, perform quantization processing on the historical fault frequency data in the analysis matrix, and output the change type of the sub-time period. Regularity analysis module: Analyze the regularity of the occurrence of the change type of the sub-time period to obtain regularity analysis data, perform quantization processing on the regularity analysis data to obtain a regularity analysis value, and compare it with the regularity analysis threshold to obtain a regularity analysis result. Among them, the regularity analysis result includes a periodic change signal or an aperiodic change signal. Prediction processing module: Based on the regularity analysis result, obtain the predicted high-frequency effective value, obtain the fault assessment within the future change fluctuation period, and complete the prediction work of the fault frequency for the future period.

[0025] Embodiment 5 Refer to Figure 3 , the embodiment of the present invention also provides a computer device 3, including: a memory 302, a processor 301, and a computer program 303 stored on the memory 302. When the computer program 303 is executed on the processor 301, it implements an integrated management and optimization system for an industrial Internet as described in any one of the above methods.

[0026] The computer device 3 may be a computing device such as a desktop computer, a notebook, a palm computer, and a cloud server. The computer device 3 may include, but is not limited to, a processor 301 and a memory 302. Those skilled in the art can understand that Figure 3 This is only an example of the computer device 3, and does not constitute a limitation on the computer device 3. It may include more or fewer components than shown in the figure, or combine some components, or different components. For example, it may also include input / output devices, network access devices, etc.

[0027] The so-called processor 301 may be a central processing unit (CPU), and this processor 301 may also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or this processor may also be any conventional processor, etc.

[0028] In some embodiments, the memory 302 may be an internal storage unit of the computer device 3, such as the hard disk or memory of the computer device 3. In other embodiments, the memory 302 may also be an external storage device of the computer device 3, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, a Flash Card, etc. equipped on the computer device 3. Further, the memory 302 may also include both the internal storage unit and the external storage device of the computer device 3. The memory 302 is used to store an operating system, application programs, a Boot Loader, data, and other programs, such as the program code of the computer program. The memory 302 may also be used to temporarily store data that has been output or is to be output.

[0029] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data and performing software simulations to obtain a formula that is closest to the actual situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0030] The above has described in detail one embodiment of the present invention, but the content described is only the preferred embodiment of the present invention and cannot be considered as limiting the scope of implementation of the present invention. Any equivalent changes and improvements made within the scope of the application of the present invention should still fall within the scope covered by the patent of the present invention.

Claims

1. A comprehensive management and optimization system for the industrial Internet, characterized in that: The following steps are involved: High-demand screening module: Analyze and process historical fault data within multiple continuous historical periods and the corresponding historical periods to obtain a historical high-fault set; The historical period is divided into several continuous historical periods of equal length, and the value of the historical fault frequency in the historical period that is greater than or equal to the rated fault frequency is recorded as the high fault frequency value; Mark the historical period corresponding to the high fault frequency value as the high fault frequency period; Type analysis module: constructs an analysis matrix for the high-frequency fault periods and high-fault frequency values ​​in the historical high-fault set, processes the historical fault frequency data in the analysis matrix, and outputs the sub-period change type; Regularity analysis module: processes the regularity of stable sub-periods and fluctuating sub-periods in multiple fault high-frequency periods to obtain periodic change signals or non-periodic change signals based on industrial equipment faults; Prediction processing module: Based on periodic change signals and non-periodic change signals, the predicted high-frequency effective values ​​under different periodic changes are obtained to complete the prediction of industrial equipment failures in future periods.

2. According to the comprehensive management and optimization system of the industrial Internet according to claim 1, it is characterized in that: The method for obtaining the historical high fault collection is as follows: Sort multiple historical periods according to the time sequence of the maintenance and transportation logs and form a historical set T; Based on any historical period in the historical set T, multiple high-frequency fault periods are integrated and counted in the time sequence of the historical period to obtain a historical high-fault set.

3. According to the comprehensive management and optimization system of the industrial Internet according to claim 1, it is characterized in that: The analysis matrix is ​​obtained as follows: In the historical high fault set, each high fault frequency period is divided into several historical high frequency sub-periods of equal time length; The historical high-frequency sub-periods divided by each fault high-frequency period are sorted by matrix to obtain the analysis matrix , where the row data in the analysis matrix R represents the historical high-frequency sub-periods divided within each fault high-frequency period, the column data in the analysis matrix R represents the historical high-frequency sub-periods divided within different fault high-frequency periods, n represents the total number of historical high-frequency sub-periods within each fault high-frequency period, and m represents the total number of fault high-frequency periods.

4. According to claim 3, the comprehensive management and optimization system of the industrial Internet is characterized in that: The historical fault frequency data in the analysis matrix is ​​quantified, and the process is as follows: Extract the column data in the analysis matrix R and get the set ; Get The historical fault frequencies corresponding to the two endpoints of the historical high-frequency sub-period are added and averaged to obtain the unit change mean; Iterating over a collection All historical high-frequency sub-periods within The processing method of historical high-frequency sub-periods is to collect Until the acquisition of the unit change mean corresponding to all historical high-frequency sub-periods is completed; The variance of all unit change means is calculated to obtain the change stability value.

5. According to claim 3, the comprehensive management and optimization system of the industrial Internet is characterized in that: The historical fault frequency data in the analysis matrix is ​​quantified, and the process is as follows: Any combination Combine the adjacent historical high-frequency sub-periods within the period and sort them according to the time series to obtain the element analysis group; Extract the earliest historical high-frequency sub-period in the element analysis group, and obtain the historical fault frequencies corresponding to the two endpoints of the earliest historical high-frequency sub-period, substitute the historical fault frequencies corresponding to the two endpoints of the earliest historical high-frequency sub-period into the two-dimensional coordinate system, and obtain the earliest period change line; Based on the change line of the previous period, the corresponding slope is calculated by the slope formula ; Similarly, extract the later historical high-frequency sub-periods within the element analysis group and calculate the corresponding slope using the slope formula. ; The slope With slope Make a difference, take the absolute value, and get the combined slope difference; The combined slope differences corresponding to all element analysis groups are added and averaged to obtain the trend similarity value.

6. According to claim 5, the comprehensive management and optimization system of the industrial Internet is characterized in that: The output is the sub-period change type, the process is as follows: Add the trend similarity value and the change stability value to obtain the sub-period comparison value; If the sub-period comparison value is greater than the sub-period comparison threshold, a period fluctuation signal is generated, and the historical high-frequency sub-period corresponding to the generated period fluctuation signal is marked as a change fluctuation sub-period; If the sub-period comparison value is less than or equal to the sub-period comparison threshold, a period stability signal is generated, and the historical high-frequency sub-period corresponding to the generated period stability signal is marked as a change-stable sub-period.

7. The comprehensive management and optimization system of the industrial Internet according to claim 1 is characterized in that: The method of obtaining regularity analysis data is as follows: Randomly select a high-frequency fault period; All stable sub-periods and fluctuating sub-periods in the fault high-frequency period are sorted according to the time sequence to obtain a sub-period sequence; Obtain the number of change fluctuation sub-periods between adjacent change stable sub-periods, add and average them, and calculate the ratio with the total number of high demand sub-periods in the fault high frequency period to obtain the number of change intervals; Combining adjacent high-frequency fault periods into a regularity analysis group, and sorting the adjacent high-frequency fault periods in the regularity analysis group according to the time sequence before and after, to obtain multiple regularity analysis groups; Based on multiple groups of regularity analysis groups, the fault frequency interval characterization value is obtained by Euclidean distance calculation .

8. The comprehensive management and optimization system of the industrial Internet according to claim 7 is characterized in that: The method of obtaining regularity analysis data is as follows: In the sub-period sequence, the high fault frequency values ​​corresponding to the adjacent stable sub-periods are subtracted and the absolute values ​​are taken to obtain the adjacent unit change values; Extract the maximum adjacent unit change value and the minimum adjacent unit change value, and perform the difference to obtain the adjacent unit change value range; The maximum adjacent unit change value and the minimum adjacent unit change value are added and averaged, and the ratio is calculated with the adjacent unit change value range to obtain a stable change value; The high fault frequency characterization value is obtained by calculating the stable change value through Euclidean distance calculation .

9. The comprehensive management and optimization system of the industrial Internet according to claim 8 is characterized in that: The periodic change identification process is: The fault frequency interval characterization value With high fault frequency characterization value Perform addition and summation to obtain the regularity analysis value; If the regularity analysis value is greater than the regularity analysis threshold, a non-periodic change signal is generated; If the regularity analysis value is less than or equal to the regularity analysis threshold, a periodic change signal is generated.

10. The comprehensive management and optimization system of industrial Internet according to claim 1, characterized in that: The method for obtaining the predicted high-frequency effective value is as follows: Based on the periodic change signal, if the future period is a stable change period, the unit change mean corresponding to the historical high-frequency sub-period is obtained, and the sum is added to obtain the average value to obtain the predicted high-frequency effective value; If the future period is a period of fluctuation, the moving average method is used to predict and obtain the predicted high-frequency effective value corresponding to the future period; Based on the non-periodic changing signal, the maximum high fault frequency value and the minimum high fault frequency value in the analysis matrix R are extracted, and the sum is added and the average is taken to obtain the predicted high-frequency effective value corresponding to the future period.

Images