A Fault Detection Method and System for a Mechanism Sand Vibrating Screen with Multi-Factor Comprehensive Analysis

Abstract

The present invention relates to the technical field of machine-made sand equipment fault monitoring and early warning, and in particular to a machine-made sand vibrating screen fault detection method and system with comprehensive analysis of multiple factors, the method steps comprising: obtaining a data benchmark under the working state of the vibrating screen according to a sampling period; obtaining the vibrating screen monitoring data according to a sampling period; comparing the vibrating screen working current with the vibrating screen working current sample to obtain the current difference; and / or comparing the vibrating screen feeding rate with the vibrating screen feeding rate sample to obtain the feeding rate difference, and according to the comparison result, finding the data benchmark closest to the vibration data; extracting the maximum amplitude benchmark and the vibration spectrum benchmark from the vibration data sample of the vibrating screen in the data benchmark closest to the data, and determining whether the vibrating screen monitoring data is normal according to the maximum amplitude benchmark and the vibration spectrum benchmark. The method of the present invention judges faults based on the relative stability of the vibration signal, which is more in line with the actual equipment production situation.

Description

Technical Field

[0001] The present invention relates to the technical field of fault monitoring and early warning of manufactured sand equipment, and particularly to a method and system for fault detection of a manufactured sand vibrating screen with multi-factor comprehensive analysis. Background Art

[0002] With the development of environmental protection and sustainable use of resources, the exploitation of natural sand resources has been gradually restricted. As a substitute product, manufactured sand production lines are constantly being applied and promoted. Among them, the vibrating screen, as the main screening equipment for manufactured sand quality control, has a relatively high failure rate due to continuous high-intensity vibration and complex stress conditions. If effective early fault warning can be carried out, production accidents can be reduced.

[0003] In the aspect of fault state monitoring of mechanical equipment, vibration signals and sound signals are the mainstream data analysis sources. In the patent "A Fault Diagnosis Method and System for a Vibration Equipment" (Publication No. CN107992801), the dynamic vibration signals collected by vibration sensors are analyzed to obtain feature weighted signals, and then contour feature signals are generated. Based on the nonlinear manifold learning method, the contour feature signals are dimensionally reduced to obtain a low-dimensional feature description of the vibration equipment. Finally, a classifier is used to classify the low-dimensional feature description to obtain the fault diagnosis result. In the patent "A Device Fault Monitoring Technology Based on Sound and Vibration Signals" (Publication No. 202210637988.7), after collecting sound and vibration signals, a device fault detection model based on a physics-informed neural network is established to diagnose the faults of the device to be tested.

[0004] In the above patents, it is basically limited to analyzing the vibration signals themselves. Once the characteristics of the vibration signals change, it is considered abnormal, without considering the differences in vibration signals brought about by the actual working conditions of the equipment. Therefore, in order to improve the accuracy of equipment fault monitoring, it is necessary to combine the working conditions of the equipment for multi-factor comprehensive analysis. Summary of the Invention

[0005] Aiming at the limitations in the prior art, the purpose of the present invention is to provide a method and system for fault detection of a manufactured sand vibrating screen with multi-factor comprehensive analysis, so as to obtain more accurate and stable fault detection results for equipment fault early warning in a manufactured sand plant.

[0006] A method for fault detection of a manufactured sand vibrating screen with multi-factor comprehensive analysis includes the following steps:

[0007] Obtain a data reference of the vibrating screen in the working state according to the sampling period, where the data reference includes a working current sample of the vibrating screen, a feeding rate sample of the vibrating screen, and a vibration data sample of the vibrating screen;

[0008] Obtain the monitoring data of the vibrating screen according to the sampling period, and the monitoring data includes the working current of the vibrating screen, the feeding rate of the vibrating screen, and the vibration data of the vibrating screen;

[0009] Compare the working current of the vibrating screen with the working current sample of the vibrating screen to obtain the current difference; and / or compare the feeding rate of the vibrating screen with the feeding rate sample of the vibrating screen to obtain the feeding rate difference, and according to the comparison result, find the data benchmark closest to the vibration data;

[0010] Extract the maximum amplitude benchmark and the vibration frequency spectrum benchmark from the vibration data samples of the vibrating screen in the closest data benchmark,

[0011] Calculate the average value of the amplitude difference between the maximum amplitude in the monitoring data and the maximum amplitude benchmark;

[0012] Calculate the average Frechet distance between the vibration frequency spectrum in the monitoring data and the vibration frequency spectrum benchmark; determine whether the monitoring data of the vibrating screen is normal according to the average value of the amplitude difference and the average Frechet distance.

[0013] As a preferred solution, both the working current sample of the vibrating screen and the working current of the vibrating screen are the average values of the vibrating screen current within the sampling period, and both the feeding rate sample of the vibrating screen and the feeding rate of the vibrating screen are the average values of the vibrating screen feeding rate within the sampling period.

[0014] As a preferred solution, if the amplitude difference is less than or equal to 0.1 A, and / or the feeding rate difference is less than or equal to 5 t / h, then the corresponding data benchmark is the data benchmark closest to the vibration data.

[0015] As a preferred solution, the method for calculating the average value of the amplitude difference is: calculate the average value of the maximum amplitude benchmark in the closest data benchmark, and subtract the average value of the maximum amplitude in the monitoring data from the maximum amplitude benchmark.

[0016] As a preferred solution, it further includes normalizing the average value of the amplitude difference.

[0017] As a preferred solution, determining whether the monitoring data of the vibrating screen is normal according to the average value of the amplitude difference and the average Frechet distance includes the following steps:

[0018] Calculate the probability of the closest data benchmark in the data benchmark, and calculate the relative average distance of the monitoring data according to the probability and the average value of the amplitude difference, and calculate the relative Frechet distance of the monitoring data according to the probability and the average Frechet distance;

[0019] Compare the relative average distance with a preset relative average distance warning threshold, and compare the relative Fréchet distance with a preset relative Fréchet distance warning threshold. Determine the number of times exceeding the warning threshold within n sampling periods. If the number of exceedances is greater than the warning count, the monitored data is abnormal.

[0020] As a preferred solution, if the monitored data of the vibrating screen itself deviates from the data benchmark beyond the threshold range, it is considered that the monitored data of the vibrating screen is abnormal.

[0021] Based on the same concept, a fault detection system for manufactured sand vibrating screens with multi-factor comprehensive analysis is also proposed, including a data collector, a fault monitoring module, a data storage module, and a human-machine interaction module.

[0022] The data collector is used to collect the monitored data of the vibrating screen, and the monitored data includes the working current of the vibrating screen, the feeding rate of the vibrating screen, and the vibration data of the vibrating screen.

[0023] The fault monitoring module is used for data processing and fault identification, including two working modes, a collection mode and a monitoring mode. Use any one of the above-mentioned fault detection methods for manufactured sand vibrating screens with multi-factor comprehensive analysis to determine whether the monitored data of the vibrating screen is normal.

[0024] The data storage module is used to store the monitored data.

[0025] The human-machine interaction module is used for system working mode switching, parameter setting, and viewing and confirming the fault monitoring data.

[0026] As a preferred solution, it also includes a wireless coverage module and a data interface module.

[0027] The wireless coverage module is used to achieve wireless network coverage and transmit the monitored data to the data interface module.

[0028] The data interface module transmits the monitored data to the fault monitoring module and the data storage module.

[0029] Based on the same concept, a computer medium is also proposed, on which instructions executable by a processor are stored. When the instructions are executed by the processor, the processor executes any one of the above-mentioned fault detection methods for manufactured sand vibrating screens with multi-factor comprehensive analysis.

[0030] Compared with the prior art, the beneficial effects of the present invention:

[0031] (1) The fault monitoring and early warning method of the manufactured sand vibrating screen of the present invention no longer judges faults based on the absolute stability of vibration signals, but judges faults based on the relative stability of vibration signals under real-time working conditions, which is more in line with the actual equipment production situation. (2) The fault monitoring and early warning system of the manufactured sand vibration of the present invention can continuously and automatically update the data reference library through the switching of control modes, and is more adaptable to the characteristic changes of vibration signals in the initial stage to the running-in period of the actual production equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 It is a schematic diagram of the data reference distribution in Embodiment 1 of the present invention;

[0033] Figure 2 It is a block diagram of the system structure of a manufactured sand vibrating screen fault detection system with multi-factor comprehensive analysis in Embodiment 2 of the present invention;

[0034] Figure 3 It is a flowchart of a manufactured sand vibrating screen fault detection method with multi-factor comprehensive analysis in Embodiment 2 of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0035] The present invention will be further described in detail below in conjunction with test examples and specific embodiments. However, this should not be understood that the scope of the above-mentioned subject matter of the present invention is limited to the following embodiments, and all technologies implemented based on the content of the present invention belong to the scope of the present invention.

[0036] Embodiment 1

[0037] The present invention provides a system of a manufactured sand vibrating screen fault detection system with multi-factor comprehensive analysis, including a data collector, a wireless coverage module, a data interface module, a fault monitoring module, a data storage module, and a human-computer interaction module.

[0038] The data collector consists of multiple vibration sensors, a storage battery, a wireless communication unit, and a connection cable. The vibration sensors are installed on the manufactured sand vibrating screen to collect the vibration signals of the vibrating screen. Other devices are installed near the vibrating screen to reduce the influence of equipment vibration. Among them, the storage battery provides working electrical energy for the data collector, the wireless communication unit is responsible for connecting to the wireless network provided by the wireless coverage module, and sending the collected vibration data to the data interface module. The connection cable is responsible for the power supply and communication connection between the vibration sensors, the storage battery, and the wireless communication unit.

[0039] The wireless coverage module is used to realize the wireless network coverage of the manufactured sand factory and forward the vibration data collected by the data collector to the data interface module.

[0040] The described data interface module serves as an internal and external data receiving component of the system. On the one hand, it receives the vibration data collected by the data collector forwarded by the wireless coverage module. On the other hand, it is connected to the safety PLC of the intelligent mechanism sand processing system, which is the equipment control device of the production line and has the function of collecting data such as the working status, working current, and feeding rate of the vibrating screen. The data interface module receives external data such as the working status, working current, and feeding rate of the vibrating screen obtained by the intelligent mechanism sand processing system. At the same time, it forwards the internal and external data to the fault detection module and the data storage module.

[0041] The described fault monitoring module is used for data processing and fault identification, including two working modes: the acquisition mode and the monitoring mode. In the acquisition mode, it obtains the data benchmark of the vibrating screen during normal operation in real time, processes it, and stores it in the data storage module. In the monitoring mode, it performs fault identification based on the real-time data, feeds back the abnormal results to the human-machine interaction module, and stores them in the data storage module at the same time.

[0042] The described data storage module is used to store data such as original data, benchmark data, and abnormal results for the fault monitoring module and the human-machine interaction module to read.

[0043] The described human-machine interaction module is used for system working mode switching, parameter setting, and viewing and confirming fault monitoring data.

[0044] The described wireless coverage module, data interface module, fault monitoring module, data storage module, and human-machine interaction module are connected through communication cables for data communication.

[0045] The present invention also provides a fault monitoring and early warning method for a mechanism sand vibrating screen with multi-factor comprehensive analysis. The method includes:

[0046] When the system equipment is in the acquisition mode and the acquisition time period T is set, when the vibrating screen is currently in a normal working state, it receives the current sample I(m) = {I 1 , I 2 , …, I m}, the feeding rate sample V(m) of the vibrating screen == {V 1 , V 2 , …, V m}, and the vibration data sample S(n) of the vibrating screen = {S 1 , S 2 , …, S n} in real time within the current time period T, where m and n are the number of data points collected within the time period T.

[0047] For the current sample I and the feeding rate sample V of the vibrating screen, calculate the average current Average feeding rate As the characteristics of the working state in this cycle:

[0048]

[0049] The first characteristic of the vibration data is the amplitude of the vibration. The larger the amplitude, the higher the vibration intensity. For the vibrating screen equipment, if the vibrating screen bolts are loose, it will cause the instability of the equipment structure and the increase of the amplitude. Therefore, for the vibration data sample S of the vibrating screen, first calculate the maximum peak value S max , which is used to evaluate the stability of the overall structure of the vibrating screen.

[0050] S max = max{S 1 , S 2 , …, S n}}.

[0051] Since it is relatively difficult to analyze the vibration signal in the time domain, it is considered to analyze it in the frequency domain. Therefore, through the fast Fourier transform formula, the spectrum data S[K] is obtained as follows:

[0052]

[0053] After the above calculations, As the data benchmark θ of the vibrating screen under normal operation in a time period T T , where reflects the current working state of the vibrating screen, and (S max , S[K]) are the vibration characteristics.

[0054] The system automatically collects and generates multiple data benchmarks {θ T1 , θ T2 , …, θ TN} as the judgment benchmark for whether the vibrating screen is normal during long-term operation. That is, when the vibrating screen is working, if its working state characteristics and the vibration characteristics (S max , S[K]) are close to the above data benchmarks, it is considered to be working normally; if it deviates from the above benchmarks, it is considered to be working abnormally.

[0055] After the system switches to the monitoring mode (the purpose of switching the mode is to transform the system from learning to recognition, so that the system can perform self-learning without the need for professional training, and directly switch to the monitoring mode for monitoring after learning), the system automatically collects the data of a time period T and generates the monitoring data in the monitoring mode In order to more accurately judge the abnormality of the monitoring data, data similar to the working state of the monitoring data should be selected from the data benchmarks for comparison. For example, the average current of the monitoring data The reference data with an average current between 39.9 A and 40.1 A should be selected for comparison, and the average vibrating screen feeding rate of the monitoring data should be monitored. The reference data between 15 t / h and 25 t / h should be selected. Among them, the deviation of 0.1 A and 5 t / h is the current fluctuation range I set by the system. scope and the vibrating screen feeding rate fluctuation range V scope , and the data reference is screened according to the settings to obtain a set of data references close to the current working state. The requirements for the set are as follows:

[0056]

[0057] The vibration characteristics are related to the working intensity, structure, and quality of the equipment. The working intensity or capacity of the equipment is generally represented by power, and power = voltage * current. However, the working voltage of the equipment is basically relatively stable. Therefore, current and power are basically equivalent, and this is the monitoring sampling value available for actual monitoring. So, the current value is used as one of the conditions for screening the data reference set. In addition, the parameters of the equipment quality are not directly collected, but its quality = the quality of the equipment itself + the quality of the sand and gravel. The quality of the equipment itself is basically considered unchanged. Therefore, the feeding speed is used to indirectly represent the quality of the sand and gravel. Therefore, in the present invention, current and vibrating screen feeding rate are used as the basis for screening the data reference set.

[0058] For the monitoring data and the screened data reference set compare S max,t and S[K] t respectively. According to the difference in the comparison of vibration characteristics, determine whether the vibrating screen is working normally. The comparison method is as follows:

[0059] The vibration characteristics extracted by the method of the present invention consist of two items, namely the maximum amplitude and the vibration frequency spectrum. Compare the two characteristics respectively. First, compare the first vibration characteristic, the maximum peak value S max,t of the monitoring data, calculate its average distance from the screened data reference set . The smaller the distance, the more normal the vibrating screen is working. The larger the distance, the more abnormal the vibrating screen is working. The normalization calculation method is used (the purpose is to unify the quantization standard. After normalization, the distance is between 0 and 1, which is easy to set the warning threshold). The calculation formula is as follows:

[0060] where N represents the number of , S MAX and S MIN are data references The maximum and minimum values of the maximum peak within.

[0061] Secondly, for the second vibration characteristic spectrum S[K] of the monitoring data t a comparison is made, and the Fréchet distance d f is used to describe the gap between the spectrum of the monitoring data and the selected data reference set in terms of spectrum (the smaller the distance, the higher the similarity between the two spectrum curves; the larger the distance, the lower the similarity between the two spectrum curves). The purpose of using the Fréchet distance d f to describe the gap between the spectrum of the monitoring data and the selected data reference set in terms of spectrum is as follows: 1) The features in the spectrum are for local feature recognition. The key is to identify the gap between the peak part in the vibration characteristic spectrum and the data reference. If the average value is used to calculate the gap, the features of the peak part will be submerged, which is not conducive to the screening of abnormal data. 2) The simple Euclidean distance should not be used because the Euclidean distance usually calculates the distance between points, and it cannot reflect the gap between the data of the peak part in the spectrum and the data reference. Through adaptive selection, finally the Fréchet distance is used to describe the gap between the spectrum of the monitoring data and the selected data reference set in terms of spectrum This can not only reflect the gap in thousands of speeds but also retain the local features (especially peak features) in the spectrum. The average Fréchet distance is calculated, and the calculation formula is as follows:

[0062] where N represents the number of , represents the Fréchet distance between S[K] t and the i-th in terms of spectrum.

[0063] Finally, since under different working conditions, the selected data reference is only a part of the data reference. If the current working condition itself is a small-probability state, then it is an anomaly in itself. As Figure 1 shown, if the data reference is selected within the range of 0.1 A and 5 t / h, then there is only 1 point in the selected data reference. Therefore, the distance between the monitoring value and the characteristic value of the data reference may be very close, but the distance of 1 point cannot fully represent its characteristics. So, probability is used to amplify the distance.

[0064] The working state of the data benchmark shows that 90% of the average current is between 40A and 45A. However, the current of the currently monitored data is only 39A. Therefore, only a small part of the data benchmark is selected. The average distance of the maximum amplitude and the average Fréchet distance of the spectrum calculated by comparison are also values under small probability events, and they cannot fully judge whether the vibrating screen is working properly. Thus, the selected data benchmark set Incorporate the probability of the entire data benchmark into the evaluation parameters for whether the vibrating screen is working properly.

[0065] For the calculated selected data benchmark set In the entire data benchmark The probability in Then, for the average distance of the calculated maximum peak And the average Fréchet distance of the spectrum Calculate the relative average distance And the relative Fréchet distance The calculation formula is as follows:

[0066]

[0067] The data benchmark is a sample under various working states. Under a certain working state, its samples will have a certain concentrated distribution area. If there are only a small number of sample distributions near the monitored value, although the distance between this value and these small numbers of samples is very close, it itself indicates a relatively abnormal event. Therefore, a distance compensation is performed using probability.

[0068] The system sets the warning threshold for the relative average distance The warning threshold for the relative Fréchet distance The warning time statistical time period range n*T and the over-limit warning times M. When within the time of n*T, the Or Then, it accumulates 1 over-limit. When the cumulative over-limit times ≥ M, a fault warning for the vibrating screen is issued.

[0069] Embodiment 2

[0070] As Figure 2 shown, the present invention provides a fault monitoring and warning system for a multi-factor comprehensive analysis of a manufactured sand vibrating screen, including a data collector, a wireless coverage module, a data interface module, a fault monitoring module, a data storage module, and a human-computer interaction module.

[0071] The data collector consists of three vibration sensors, batteries, wireless communication units, and connecting cables. One vibration sensor is installed on the vibrating screen motor, and two vibration sensors are installed on both sides of the vibrating screen, one high and one low. The vibration sensor uses a frequency of 2KHz; the battery and wireless communication unit are installed below the outer side of the safety fence to prevent the vibration of the vibrating screen from affecting the wireless data transmission, and to prevent people from walking and touching the equipment. Among them, the vibration sensor is used to collect vibration signals in real time, the battery provides working power for the data collector, and the wireless communication unit is responsible for connecting to the wireless network provided by the wireless coverage module. It is necessary to set up automatic connection to the specified wireless network to meet the automatic connection after power-on, and send the collected vibration data to the data interface module. The connecting cable is responsible for the power supply and communication connection between the vibration sensor, battery, and wireless communication unit.

[0072] The wireless coverage module is used to achieve wireless network coverage of the machine-made sand plant. It adopts wifi technology to receive the vibration signal sent by the data collector in real time and transmit it to the data interface module through the wired network.

[0073] The data interface module serves as an internal and external data receiving component of the system. It receives the collected vibration signals in real time internally. It is connected to the safety PLC of the artificial sand intelligent processing system externally to receive external data such as the vibrating screen working status, working current, and vibrating screen feeding rate in real time. The above external data is sent regularly by the artificial sand intelligent processing system at a transmission frequency of 2s / time. The data interface module forwards the above data to the fault monitoring module and the data storage module in real time.

[0074] The fault monitoring module is used for data processing and fault identification. It includes two working modes, acquisition mode and monitoring mode. The acquisition mode calculates the data benchmark for the normal operation of the vibrating screen in real time and stores it in the data storage module after processing. The monitoring mode identifies faults based on real-time data, system parameters of the data storage module and data benchmark, and feeds back abnormal results to the human-computer interaction module and stores them in the data storage module at the same time. The fault monitoring module does not calculate at any time. It will also make judgments based on the working status of the vibrating screen. It will only calculate when the vibrating screen is in operation. When the vibrating screen is in other states (such as stopped, ready, etc.), the real-time received data will be discarded.

[0075] The data storage module is used for data storage of the system. It can store the data benchmark in the acquisition mode, the over-limit abnormality and warning records in the monitoring mode, the system setting parameters and the original internal and external interface data in the database for reading by the fault monitoring module and the human-computer interaction module.

[0076] The human-computer interaction module is used for system working mode switching, parameter setting, and viewing and confirming fault monitoring data.

[0077] When the fault monitoring and early warning system of the manufactured sand vibrating screen of the present invention is specifically used, it operates according to the following steps:

[0078] Step 1: After the manufactured sand production line operates normally, check whether the current vibrating screen is working properly. After confirming that it is normal, proceed to Step 2;

[0079] Step 2: First, set the basic parameters for the system operation, including setting the data acquisition period to 30s, the control mode to the acquisition mode, the current fluctuation range to 0.1A, and the feeding rate fluctuation range of the vibrating screen to 5t / h.

[0080] Step 3: Start the system to collect data and continuously generate reference data. This stage lasts for 2 weeks to collect sufficient reference data.

[0081] Step 4: Switch the system control mode to the monitoring mode and initially set the early warning threshold for the relative average distance to 0.5, and the early warning threshold for the relative Fréchet distance to 0.5. The threshold is set relatively small at this time, and the purpose is to make the system give early warning prompts.

[0082] Step 5: The system runs continuously for one week in this stage. Select the maximum relative average distance and the early warning threshold of the maximum Fréchet distance from the early warning information Then consider a 10% margin and set the early warning threshold for the relative average distance and the early warning threshold for the relative Fréchet distance

[0083] Step 6: After the system debugging is completed, it can officially enter the monitoring mode. If there are frequent fault early warnings in the later stage, the equipment needs to be inspected, repaired, and restored to its normal state; if it still continues to alarm after the inspection and repair is normal, the operations in Steps 1 to 5 can be repeated, and the duration of Steps 3 and 5 can be considered to be shortened to quickly restore the system.

[0084] As Figure 2 shown, the present invention provides a fault monitoring and early warning method for a manufactured sand vibrating screen with multi-factor comprehensive analysis. This method is implemented by the fault monitoring module of the system, and the specific implementation is as follows:

[0085] S1: The system sets parameters such as the acquisition period of 30s, the current fluctuation range of 0.1A, and the feeding rate fluctuation range of the vibrating screen of 5t / h. Manually check and confirm that the current equipment is normal, switch the system working mode to the acquisition template, and the fault monitoring module obtains the system setting parameter information in real time.

[0086] S2: During the real-time operation of the fault monitoring module, it receives internal and external interface data in real time. For the external interface data, when real-time current, vibrating screen feeding rate, and vibration signals for a cycle of 30 s are collected and received, it is necessary to judge according to the working state of the vibrating screen obtained from the external interface to confirm the next process.

[0087] S3: If the vibrating screen operates normally continuously during the acquisition period, execute S5; if not, execute S4.

[0088] S4: Discard the acquisition data of this cycle, return to S2, and continue to collect data for the next cycle.

[0089] S5: The fault monitoring module receives the current samples I(m) = {I 1 , I 2 , …, I m} and the vibrating screen feeding rate samples V(m) == {V 1 , V 2 , …, V m} within the cycle time of 30 s in real time. According to the data transmission frequency of 2 s / time, there are approximately 15 data values, and calculate the average current and the average vibrating screen feeding rate . The calculation formulas are as follows:

[0090]

[0091] The fault monitoring module receives the vibration signal data samples S(n) = {S 1 , S 2 , …, S n} within the cycle time of 30 s in real time. According to the sampling frequency of the vibration sensor, there are approximately 60,000 data, calculate the maximum peak value S max of the vibration signal and obtain the frequency spectrum data S[K] through the fast Fourier transform formula. The calculation formulas are as follows:

[0092] S max = max{S 1 , S 2 , …, S n}

[0093]

[0094] S6: Judge the current system working mode. If it is in the monitoring mode, execute S8; if the current system is in the acquisition mode, execute S7.

[0095] S7: Take the data after calculation and processing as the data reference and store it in the data reference library Then return to S2 to continue collecting data for the next cycle. The collection of the data reference needs to last for a sufficient time so that the system can collect enough data references.

[0096] S8: According to the parameter current fluctuation range I set in S1 scope = 0.1A, the vibrating screen feed rate fluctuation range V scope = 5t / h, the average current of the data collected in this cycle and the average vibrating screen feed rate Select the data reference set in the current working state from the data references The selection requirements are as follows:

[0097]

[0098] S9: Calculate the maximum peak value S of the data collected in this cycle max,t , calculate its average distance from the selected data reference set of using a normalized calculation method. The calculation formula is as follows: where N represents the number of

[0099] and S and S MAX and S MIN are the maximum and minimum values of the maximum peak values in the data reference .

[0100] S10: Use the Fréchet distance d t to describe the gap between the spectrum S[K] of the monitored data and the spectrum of the selected data reference set f , and calculate the average Fréchet distance. The calculation formula is as follows: where N represents the number of ,

[0101] and represents the Fréchet distance between S[K] and the i-th t .

[0102] S11: Calculate the probability of the data reference selected in the current working state in the entire data reference , and then calculate the relative average distance of the average distance of the calculated maximum peak value and the average Fréchet distance of the spectrum and the relative Fréchet distance The calculation formula is as follows: ​

[0103]

[0104] In S1, the system needs to set in advance the warning threshold of the average distance of the maximum peak relative to the average distance Warning threshold relative to the Fréchet distance Warning time statistical time period range n*T and the number of over-limit warnings M, where and It is determined according to the aforementioned system operation steps 1 to 5, and the warning time statistical period can be set to 5*T, that is, 150 s, and the number of over-limit warnings can be set to 3 times.

[0105] S12: Determine the relative average distance of this period or the relative Fréchet distance Whether it exceeds the limit. If any value exceeds the limit, it is determined that the data of this period exceeds the limit. If it exceeds the limit, the over-limit count of this period is counted as 1, otherwise it is counted as 0.

[0106] S13: Calculate the cumulative over-limit count within the previous 5 periods from the current time. If it is greater than or equal to 3 times, execute S14, otherwise directly return to S2 to continue collecting data for the next period.

[0107] S14: The system gives a vibration screen fault warning and displays the latest calculation result for maintenance personnel to view, and then returns to S2 to continue collecting data for the next period.

[0108] The above shows and describes the basic principles, main features and advantages of the present invention. For those skilled in the art, it is obvious that the present invention can be implemented in other specific forms without departing from the spirit or basic features of the present invention. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the claims are intended to be embraced within the present invention. Any reference signs in the claims should not be regarded as limiting the claims involved.

[0109] In addition, it should be understood that although this specification is described according to embodiments, not every embodiment only contains an independent technical solution. This narrative way of the specification is only for clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in the embodiments can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A fault detection method for a vibratory screen of manufactured sand with multi-factor comprehensive analysis, characterized in that, it includes the following steps: Obtain the data benchmark under the working state of the vibratory screen according to the sampling period, and the data benchmark includes the working current sample of the vibratory screen, the feeding rate sample of the vibratory screen, and the vibration data sample of the vibratory screen; Obtain the monitoring data of the vibratory screen according to the sampling period, and the monitoring data includes the working current of the vibratory screen, the feeding rate of the vibratory screen, and the vibration data of the vibratory screen; Compare the working current of the vibratory screen with the working current sample of the vibratory screen to obtain the current difference; and / or compare the feeding rate of the vibratory screen with the feeding rate sample of the vibratory screen to obtain the feeding rate difference, and according to the comparison result, find the data benchmark closest to the vibration data; Extract the maximum amplitude benchmark and the vibration frequency spectrum benchmark from the vibration data sample of the vibratory screen in the closest data benchmark; Calculate the average value of the amplitude difference between the maximum amplitude in the monitoring data and the maximum amplitude benchmark; Calculate the average Fréchet distance between the vibration frequency spectrum in the monitoring data and the vibration frequency spectrum benchmark; Determine whether the monitoring data of the vibratory screen is normal according to the average value of the amplitude difference and the average Fréchet distance; Determine whether the monitoring data of the vibratory screen is normal according to the average value of the amplitude difference and the average Fréchet distance, including the following steps: Calculate the probability of the closest data benchmark in the data benchmark, and calculate the relative average distance of the monitoring data according to the probability and the average value of the amplitude difference, and calculate the relative Fréchet distance of the monitoring data according to the probability and the average Fréchet distance; Compare the relative average distance with the preset relative average distance warning threshold, compare the relative Fréchet distance with the preset relative Fréchet distance warning threshold, and judge the number of times exceeding the warning threshold within n sampling periods. If the number of times exceeding is greater than the warning number, the monitoring data is abnormal; If the monitoring data of the vibratory screen itself deviates from the data benchmark beyond the threshold range, it is considered that the monitoring data of the vibratory screen is abnormal.

2. The fault detection method for a vibratory screen of manufactured sand with multi-factor comprehensive analysis according to claim 1, characterized in that, Both the working current sample of the vibratory screen and the working current of the vibratory screen are the average values of the vibratory screen current within the sampling period, and both the feeding rate sample of the vibratory screen and the feeding rate of the vibratory screen are the average values of the vibratory screen feeding rate within the sampling period.

3. The fault detection method for a vibratory screen of manufactured sand with multi-factor comprehensive analysis according to claim 1, characterized in that, If the amplitude difference is less than or equal to 0.1 A, and / or the feeding rate difference is less than or equal to 5 t / h, the corresponding data benchmark is the data benchmark closest to the vibration data.

4. The fault detection method for a vibratory screen of manufactured sand with multi-factor comprehensive analysis according to claim 1, characterized in that, The calculation method of the average value of the amplitude difference is: calculate the average value of the maximum amplitude benchmark in the closest data benchmark, and subtract the average value of the maximum amplitude in the monitoring data from the maximum amplitude benchmark.

5. A method for detecting faults of a manufactured sand vibrating screen with multi-factor comprehensive analysis as described in claim 4, characterized in that, it further includes normalizing the average value of the amplitude difference.

6. A system for detecting faults of a manufactured sand vibrating screen with multi-factor comprehensive analysis, characterized in that, it includes a data collector, a fault monitoring module, a data storage module and a human-computer interaction module, the data collector is used to collect the monitoring data of the vibrating screen, and the monitoring data includes the working current of the vibrating screen, the feeding rate of the vibrating screen and the vibration data of the vibrating screen; the fault monitoring module is used for data processing and fault identification, and includes two working modes, a collection mode and a monitoring mode, and uses a method for detecting faults of a manufactured sand vibrating screen with multi-factor comprehensive analysis described in any one of claims 1-5 to judge whether the monitoring data of the vibrating screen is normal; the data storage module is used to store the monitoring data; the human-computer interaction module is used for switching the system working mode, setting parameters, and viewing and confirming the fault monitoring data.

7. A system for detecting faults of a manufactured sand vibrating screen with multi-factor comprehensive analysis as described in claim 6, characterized in that, it further includes a wireless coverage module and a data interface module, the wireless coverage module is used to achieve wireless network coverage and transmit the monitoring data to the data interface module; the data interface module transmits the monitoring data to the fault monitoring module and the data storage module.

8. A computer medium, characterized in that, it stores instructions executable by a processor, and when the instructions are executed by the processor, the processor executes a method for detecting faults of a manufactured sand vibrating screen with multi-factor comprehensive analysis described in any one of claims 1-5.

Images