Current linkage early warning method used under coal mine underground multi-motor coupling effect

Abstract

The invention belongs to the technical field of current safety production, and provides a current linkage early warning method under a coal mine underground multi-motor coupling effect, which comprises the following steps of: setting a sensor network, and collecting a historical current sequence and a historical working condition sequence of each motor, marking abnormal working conditions and corresponding omen working conditions in a historical working condition sequence of each motor, setting a monitoring window, judging whether the abnormal working conditions or the omen working conditions exist in the monitoring window or not, if not, triggering current prediction, and constructing a fusion prediction model which comprises a time sequence current prediction model and a time sequence working condition prediction model, and obtaining a motor group coupling model established by basic information of each motor, predicting each motor through a fusion prediction model to obtain a predicted current sequence of each motor, analyzing a historical current sequence in the monitoring window and the predicted current sequence obtained through prediction, analyzing current abnormity of the motor, and performing early warning.

Description

Technical Field

[0001] The invention belongs to the technical field of current production safety, and in particular relates to a current linkage early warning method for multi-motor coupling effects in underground coal mines. Background Art

[0002] In coal mining operations, the coordinated operation of multiple motors underground is a key link in ensuring the smooth progress of the production process. Many motors jointly drive belt transportation, ventilation, drainage and other systems. However, with the continuous increase in the depth and scale of coal mining, the operating environment of underground motors has become increasingly complex. Harsh conditions such as high temperature, humidity, and dust not only affect motor performance, but also significantly increase the risk of motor failure. At the same time, the interactions such as electromagnetic coupling and mechanical coupling between multiple motors cause one motor to malfunction, which can easily trigger a chain reaction, affecting the stable operation of the entire system and even causing serious safety accidents. Traditional motor monitoring systems often rely on only a single type of sensor to collect data, such as monitoring only current or temperature, which cannot fully reflect the operating status of the motor. Due to the complex causes of motor failures in coal mines, single parameter monitoring is difficult to capture potential fault signals in the early stages. As a result, many faults are not discovered until they have developed to a serious stage, delaying maintenance opportunities, causing production interruptions and economic losses. On the other hand, existing early warning methods mostly use fixed thresholds to judge abnormalities, which lack adaptability to dynamic changes in motor operating conditions. Coal mine underground operating scenes are changeable, and motor loads and environmental conditions may change at any time. Fixed thresholds cannot accurately distinguish between normal fluctuations and true abnormal conditions, and are prone to false alarms or missed alarms. For example, when underground equipment is started or the load changes suddenly, parameters such as current will fluctuate briefly, and traditional threshold judgments may misjudge them as abnormalities; and when the motor has a minor fault and the parameter change does not reach the fixed threshold, it may not be possible to issue an early warning in time. In addition, the existing technology has failed to fully consider the impact of the coupling effect between multiple motors on the operating status of the motors. In a multi-motor system, the interaction between the motors will cause the changes in parameters such as current and speed to present complex nonlinear relationships. Traditional methods are difficult to accurately predict the operating trends of the motors under this coupling effect, and it is impossible to take effective measures in advance to avoid the expansion and spread of faults. In view of the above problems, the present invention proposes a current linkage early warning method for multi-motor coupling effect in coal mines. Summary of the Invention

[0003] In order to make up for the deficiencies of the prior art, at least one technical problem raised in the background technology is solved.

[0004] The technical solution adopted by the present invention to solve the technical problem is: a current linkage early warning method for multi-motor coupling effect in coal mines, comprising: A sensor network is set up for the multi-motor linkage system in underground coal mines to collect the historical current sequence and historical operating condition sequence of each motor; Mark abnormal working conditions and corresponding foreboding working conditions in the historical working condition sequence of each motor, set a monitoring window, and determine whether there are abnormal working conditions or foreboding working conditions in the monitoring window. If not, trigger current prediction; If current prediction is triggered, a fusion prediction model is constructed. The fusion prediction model includes a time series current prediction model and a time series operating condition prediction model trained based on historical current sequences and historical operating condition sequences, respectively, as well as a motor group coupling model established by obtaining basic information of each motor. The fusion prediction model is used to predict each motor and obtain a predicted current sequence for each motor. Analyze the historical current sequence within the monitoring window and the predicted current sequence, analyze the motor current anomaly and issue an early warning; The historical current sequence and the historical operating condition sequence are obtained as follows: A sensor network is set up for the multi-motor linkage system in underground coal mines. The motor parameters of each motor are collected in real time according to the set data collection frequency. The time when the sensor network collects data in real time according to the data collection frequency is marked as the collection moment. At each collection moment, the three-phase stator current vector and operating condition data vector of each motor are constructed. The period between the time when the sensor network starts collecting data and the current time is marked as the historical period. For each motor, the three-phase stator current vector and the operating condition data vector collected during the historical period are integrated according to the time sequence to obtain the historical current sequence and historical operating condition sequence of each motor. The method for determining whether an abnormal operating condition or a premonitory operating condition exists is as follows: Set a monitoring window with the current time as the end point and a constant duration, and determine whether there is overlap between the monitoring window and the time period corresponding to any abnormal working condition. If so, determine that there is an abnormal working condition in the monitoring window; If it does not exist, set several omen analysis periods in the monitoring window, all of which end at the current time and have different durations, intercept the historical operating condition sequences of the motor in all omen analysis periods, obtain the omen analysis sequences of the motor, and normalize them respectively to obtain a normalized omen analysis sequence; Obtain the normalized precursor sequences of all precursor conditions. If the DTW distance between any normalized precursor analysis sequence and any normalized precursor sequence within the monitoring window is less than a distance threshold, it is determined that a precursor condition exists within the monitoring window. The abnormal operating condition is determined as follows: Obtain and calculate the mean and standard deviation of each parameter of the operating condition data vector in the historical operating condition sequence of each motor within the normal operating condition data range. Determine the dynamic abnormality threshold through data processing. If any parameter of the operating condition data vector at any acquisition moment exceeds the set dynamic abnormality threshold, assign an abnormal moment mark including the motor number to the acquisition moment. For any motor number, if abnormal time mark exists at no less than KL consecutive collection moments, the corresponding motor is determined to be in abnormal working condition, and the collection moments with consecutive abnormal time mark are demarcated as an abnormal working condition of the corresponding motor. Adjacent abnormal working conditions of the same motor with a time distance less than the preset time threshold are merged into one abnormal working condition; The method for determining the premonition condition is as follows: Obtain all the precursor sequence groups and the normalized precursor sequences they contain, process the number of normalized precursor sequences in the same precursor sequence group and the total number of abnormal sequences in the corresponding abnormal working conditions of the same type to obtain a stable ratio; If the DTW distance between any two normalized omen sequences that do not belong to the same omen sequence group meets the similarity criterion, the two omen sequence groups are classified into the same similarity set, the number of omen sequence groups in the similarity set where the omen sequence group belongs is obtained, and the data is processed with the total number of all omen sequence groups to obtain the similarity ratio; If the stability ratio is greater than the preset stability standard and the similarity ratio is less than the preset similarity standard, the omen sequence corresponding to the normalized omen sequence in the omen sequence group is judged to be the omen working condition; The method for obtaining the precursor sequence group is as follows: In the motor's historical operating condition sequence, the motor's abnormal sequence is intercepted and normalized respectively to obtain a normalized abnormal sequence and the normalized operating condition data vector. The DTW distance between any two abnormal sequences of the same motor and the corresponding normalized abnormal sequence is calculated. If the distance is less than a distance threshold, the two abnormal sequences are of the same type of abnormal operating condition. In the motor's historical operating condition sequence, the motor's precursor sequences are intercepted and normalized to obtain normalized precursor sequences. Within the same abnormal operating condition, the DTW distance between any two abnormal sequences corresponding to the normalized precursor sequences is calculated. If the distance is less than a distance threshold, the two sequences are classified into the same precursor sequence group. The predicted current sequence is obtained as follows: Obtain the time series current training set, calculate the root mean square error of the time series current prediction model and the motor group coupling model on the time series current training set, and set weights for the time series current prediction model and the motor group coupling model according to the root mean square error. The fusion prediction model performs weighted fusion of the prediction results of the time series current prediction model and the motor group coupling model based on the set weights. That is, the time series prediction current sequence and the physical prediction current sequence are weightedly fused through the calculated weights to obtain the predicted current sequence of each motor. The method for obtaining the time series prediction current sequence is as follows: Eliminate all abnormal and foreboding operating condition data vectors and three-phase stator current vectors from the historical operating condition sequence and historical current sequence. Use interpolation to supplement and normalize the historical operating condition sequence and historical current sequence to obtain a time-series operating condition training set and a time-series current training set. Obtain the monitoring window and intercept it to obtain the window operating condition sequence and window current sequence. Using a long short-term memory network, a time-series working condition prediction model and a time-series current prediction model are constructed and trained respectively using a time-series working condition training set and a time-series current training set. Combining the window working condition sequence and the window current sequence, the time-series predicted working condition sequence and the time-series predicted current sequence of the motor are obtained. The physical prediction current sequence is obtained as follows: Obtain basic information about each motor and obtain the loss parameters of each motor through statistics of historical current sequences and historical operating condition sequences. Based on the motor dynamics equations and Kirchhoff's laws, combined with the multi-motor coupling effect, establish a motor group coupling model with the input being the operating condition data vector of each motor and the output being the three-phase stator current vector of each motor. Input the time-series predicted operating condition sequence of each motor into the motor group coupling model to obtain the physically predicted current sequence of each motor. The analyzing of the abnormal current of the motor includes: Obtain a window current sequence and a predicted current sequence. The window current sequence includes the three-phase stator current vector, and the predicted current sequence includes the predicted three-phase stator current vector. Calculate the current effective value and the predicted current effective value through data processing. If the current effective value or the predicted current effective value meets the current overload abnormality condition for at least KL consecutive acquisition moments, then determine that the corresponding motor has a current overload abnormality. Otherwise, the three-phase imbalance of the motor and the predicted three-phase imbalance are calculated. If the three-phase imbalance or the predicted three-phase imbalance is greater than the national standard critical value at no less than KL consecutive acquisition moments, it is determined that the corresponding motor has a three-phase imbalance abnormality.

[0005] The beneficial effects of the present invention are as follows: 1. The present invention sets up a sensor network to comprehensively collect the historical current and operating condition sequences of each motor, and marks abnormal operating conditions and precursory operating conditions. It can accurately capture key information in motor operation, and build a fusion prediction model based on these data. It comprehensively considers the current, operating conditions and motor group coupling effects to make the prediction results more in line with reality. Compared with the traditional single monitoring method, it can detect potential current anomalies in advance, effectively avoid production interruptions caused by misjudgment, significantly improve the continuity and stability of coal mine production, and reduce the economic losses caused by motor failures.

[0006] 2. The present invention can monitor the motor operating status in real time and dynamically by setting a monitoring window and conducting analysis and early warning. When there is no abnormality or foreboding working condition in the monitoring window, current prediction is triggered, and the future current change trend of the motor is grasped in time. Once a current abnormality is found, an early warning can be issued quickly to buy valuable processing time for the staff. This helps to take maintenance measures in advance, prevent the expansion of faults, and ensure the safety of underground coal mine workers and equipment. At the same time, it improves the safety and reliability of coal mine production and reduces the probability of safety accidents. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] The present invention will be further described below with reference to the accompanying drawings.

[0008] Figure 1 This is a flowchart of the steps of a current linkage early warning method for multi-motor coupling effect in coal mines according to an embodiment of the present invention; Figure 2 This is a flow chart of the steps for obtaining a predicted current sequence in a current linkage early warning method under the coupling effect of multiple motors in underground coal mines, as described in an embodiment of the present invention. DETAILED DESCRIPTION

[0009] In order to make the technical means, creative features, objectives and effects achieved by the present invention easier to understand, the present invention is further described below in conjunction with specific implementation methods.

[0010] Example 1 See also Figure 1 As shown, the current linkage early warning method for multi-motor coupling effect in coal mines according to an embodiment of the present invention includes the following steps: S1: Set up a sensor network to collect the historical current sequence and historical operating condition sequence of each motor; A sensor network is set up for the multi-motor linkage system in an underground coal mine. The sensor network includes current sensors, vibration sensors, temperature sensors, and speed sensors. The data collection frequency is determined based on the working characteristics and monitoring requirements of the multi-motor linkage system in the underground coal mine. The motor parameters of each motor in the multi-motor linkage system in the underground coal mine are collected in real time through the sensor network according to the data collection frequency. The motor parameters include three-phase stator current, vibration acceleration of the bearing seat, winding temperature, bearing temperature, rotor speed, and input voltage. The time when the sensor network collects the motor parameters of each motor in real time according to the data collection frequency is marked as the collection time. Among them, the current sensors are connected in series to the three-phase circuits of each motor distribution cabinet in the multi-motor linkage system in the coal mine to ensure the consistency of the current direction. The current sensors use closed-loop high-precision explosion-proof Hall current sensors to collect the three-phase stator current of each motor in real time. 、 as well as , where m represents the number of the motor in the multi-motor linkage system in the coal mine, and t represents the collection time of the sensor network for data collection; Among them, the vibration sensors are rigidly fixed to the radial bearing area of ​​each motor bearing seat in the multi-motor linkage system in the coal mine through special bolts. The vibration sensors use wide-band response accelerometers to capture the vibration acceleration of each motor bearing seat. ; Among them, temperature sensors are pre-buried in the stator windings of each motor in the multi-motor linkage system in the coal mine, and surface mount temperature sensors are installed at the bearings of each motor to synchronously collect the winding temperature. and bearing temperature ; The speed sensor uses a photoelectric / magnetoelectric dual-mode probe. The distance between the photoelectric probe and the reflective code disk is (1.0±0.2) mm, and the air gap between the magnetoelectric probe and the reflective code disk is (1.5±0.5) mm. The speed sensor is used to monitor the rotor speed of each motor in the multi-motor linkage system in the coal mine in real time. ; Among them, the voltage sensor uses a high-precision explosion-proof voltage transformer connected in parallel to the power supply line of each motor in the multi-motor linkage system in the coal mine. The ratio is adapted to the rated voltage to collect the input voltage of each motor. ; For the motor parameters of each motor in the multi-motor linkage system of the coal mine, the three-phase stator current vector of each motor is constructed at each acquisition moment. and the working condition data vector ; ; ; Edge pre-processing and hybrid networking architecture are used to achieve efficient data transmission, eliminate random noise and grid interference, and use GPS / IRIG-B dual-mode timing modules to synchronize the motor parameter time scales of each motor; The period between the time when the sensor network starts collecting data and the current time is marked as the historical period. For each motor, the three-phase stator current vector and the operating condition data vector collected during the historical period are integrated according to the time sequence to obtain the historical current sequence and historical operating condition sequence of each motor. It should be noted that the purpose of this step is to build a sensor network containing multiple sensors, determine the acquisition frequency according to the working characteristics of the coal mine motor, obtain the motor's three-phase stator current, vibration acceleration, temperature, speed and voltage parameters in real time, build the three-phase stator current vector and operating condition data vector of each motor, and optimize data transmission and synchronize time scales. Finally, the historical current series and historical operating condition series are integrated to provide basic data for subsequent analysis. Edge preprocessing and hybrid networking architecture are used to eliminate noise and interference and improve data quality. Time scale synchronization ensures data time consistency, which facilitates subsequent timing analysis. S2: Mark abnormal working conditions and corresponding foreboding working conditions in the historical working condition sequence of each motor, set a monitoring window, and determine whether there are abnormal working conditions or foreboding working conditions in the monitoring window. If not, trigger current prediction; In the multi-motor linkage system of a coal mine at the current moment, the historical operating condition sequence of each motor is obtained. Based on the normal operating condition data range screened by preliminary rules and manually marked, the mean value of the nth parameter of the operating condition data vector in the historical operating condition sequence of motor number m is calculated under the normal operating condition data range. and standard deviation , set the dynamic anomaly threshold to: ; Where k is a manually preset empirical coefficient. For each motor, in the historical operating condition sequence, if any parameter of the operating condition data vector at any acquisition moment exceeds the preset dynamic abnormality threshold, the motor is determined to have a parameter abnormality at that acquisition moment, and an abnormal moment mark including the motor number is assigned to the acquisition moment. To avoid misjudgment due to individual noise data, a continuous judgment mechanism is adopted. For any motor number, if there are abnormal time mark at no less than KL consecutive collection moments, the corresponding motor is determined to be in an abnormal working condition. Otherwise, the abnormal time mark of the corresponding motor number at the collection moment is cleared, and the collection moments with continuous abnormal time mark of the same motor number are demarcated as an abnormal working condition of the corresponding motor. For any motor, adjacent abnormal working conditions with a time distance less than a preset time threshold are merged into one abnormal working condition. Based on any motor, all abnormal operating conditions are extracted from the motor's historical operating condition sequence, and the motor's abnormal sequence is intercepted. Within each abnormal sequence, each parameter of the operating condition data vector is normalized using the minimum-maximum normalization method to obtain a normalized abnormal sequence and a normalized operating condition data vector within the normalized abnormal sequence; For any two abnormal sequences of the same motor, the DTW distance between the two corresponding normalized abnormal sequences is calculated using the dynamic time warping (DTW) algorithm to analyze the similarity of the change trends between the two abnormal sequences. Specifically, an M×N distance matrix D is constructed, where M and N are the lengths of the two normalized anomaly sequences, respectively. The matrix element D(i,j) represents the Euclidean distance between the i-th and j-th normalized operating condition data vectors in the two normalized anomaly sequences. The shortest path from the upper left corner to the lower right corner of the distance matrix is ​​calculated using a dynamic programming method. The cumulative distance of the shortest path is the DTW distance of the two normalized anomaly sequences. The smaller the DTW distance, the more similar the change trends of the two normalized anomaly sequences are. The abnormal sequences corresponding to the normalized abnormal sequences of the same motor whose DTW distance is less than the distance threshold are divided into the same type of abnormal working conditions; Set a precursor collection period for each abnormal sequence. The end of the precursor collection period is the starting point of the corresponding abnormal sequence. The duration of the precursor collection period is determined by the duration of the abnormal sequence. For the same motor, the benchmark duration is set for the precursor collection period of the first abnormal sequence in the historical period. The ratio of the duration of all abnormal sequences of the same motor to the duration of the corresponding precursor collection period is the same. In the historical working condition sequence of the motor, the historical working condition sequence within all the prognostic collection periods of the motor is intercepted to obtain the prognostic sequence of the motor. In each prognostic sequence, each parameter of the working condition data vector is normalized using the minimum-maximum normalization method to obtain a normalized prognostic sequence. Similarly, within the same type of abnormal working conditions, obtain the normalized precursor sequences corresponding to any two abnormal sequences, calculate the DTW distance between the two normalized precursor sequences using the dynamic time warping algorithm, and classify the two normalized precursor sequences with a DTW distance less than the distance threshold into the same precursor sequence group. If there is a precursor sequence that is not classified into any precursor sequence group, the precursor sequence itself is a precursor sequence group. Obtain the number of normalized precursor sequences within the same precursor sequence group and perform ratio processing on it with the total number of abnormal sequences in the corresponding abnormal working conditions of the same type to obtain the stable ratio; For two different omen sequence groups, if the DTW distance between any two normalized omen sequences that do not belong to the same omen sequence group is less than the distance threshold, the two omen sequence groups are judged to be similar and are classified into the same similarity set. If any omen sequence group is not classified into any similarity set, the omen sequence group itself is a similarity set. Obtain the number of omen sequence groups in the similarity set where any omen sequence group is located, and perform ratio processing on it with the total number of all omen sequence groups to obtain the similarity ratio; If the stability ratio is greater than the preset stability standard and the similarity ratio is less than the preset similarity standard, the precursor sequence corresponding to the normalized precursor sequence in the precursor sequence group is judged to be a precursor condition of the same type of abnormal condition; It should be noted that, based on the preset ratio threshold, the same type of abnormal working condition may have multiple precursor working conditions, or may not find a precursor working condition; Set a monitoring window, which is a period of time with the current time as the end point and a constant length. The monitoring window slides over time to determine whether there is overlap between the monitoring window and the period corresponding to any abnormal sequence. If so, it is determined that there is an abnormal operating condition within the monitoring window; If it does not exist, set several omen analysis periods in the monitoring window. The omen analysis period ends at the current time. The duration of all omen analysis periods in the monitoring window is different. Among them, the longest omen analysis period is the same as the duration of the monitoring window. In the historical operating condition sequence of the motor, the historical operating condition sequence of the motor within all the prognostic analysis periods is intercepted to obtain the prognostic analysis sequence of the motor. In each prognostic analysis sequence, each parameter of the operating condition data vector is normalized using the minimum-maximum normalization method to obtain a normalized prognostic analysis sequence. If the DTW distance between any normalized precursor analysis sequence and any normalized precursor sequence of a precursor condition in the monitoring window is less than the distance threshold, it is determined that a precursor condition occurs in the monitoring window; For any motor, if an abnormal operating condition or a foreboding operating condition occurs within the monitoring window, an alarm is generated and the motor number is sent to the administrator terminal; If all motors in the coal mine underground multi-motor linkage system have no abnormal working conditions and premonitory working conditions within the monitoring window, it is judged that the coal mine underground multi-motor linkage system is operating normally, and the current prediction is triggered; It should be noted that the purpose of this step is to mark abnormal working conditions and their precursory working conditions through statistical analysis and clustering algorithms based on historical working condition data, and introduce the dynamic time warping (DTW) algorithm to perform cluster analysis on abnormal working conditions and precursory working conditions. This breaks through the limitations of traditional single threshold judgment and can capture complex changing trends. Predictive working conditions are judged based on stable ratios and similarity ratios, and a more scientific abnormal precursor judgment standard is established. A monitoring window is set to determine in real time whether there are abnormal working conditions or precursory working conditions. If they do not appear, current prediction is triggered. The mining of precursory working conditions realizes early warning of faults, buying time for preventive maintenance. The monitoring window mechanism ensures real-time monitoring, timely detection of potential risks, and early identification of abnormalities and early warning triggering decisions. S3: If current prediction is triggered, a fusion prediction model is constructed. The fusion prediction model includes a time series current prediction model and a time series operating condition prediction model, which are trained based on historical current sequences and historical operating condition sequences, respectively, as well as a motor group coupling model established by obtaining basic information about each motor. The fusion prediction model is used to predict each motor and obtain a predicted current sequence for each motor. like Figure 2 As shown, the specific steps for obtaining the predicted current sequence are as follows: If current prediction is triggered, a fusion prediction model of the multi-motor linkage system in the coal mine is constructed. The fusion prediction model includes a time series current prediction model, a time series operating condition prediction model, and a motor group coupling model. Specifically, the historical current sequence of each motor is obtained, and all abnormal operating conditions and three-phase stator current vectors within the precursor collection period are removed from the historical current sequence. The historical current sequence is supplemented by interpolation and normalized to obtain a time-series current training set. The time-series current training set within the monitoring window is intercepted to obtain a window current sequence. Similarly, the historical operating condition sequence of each motor is obtained, and all abnormal operating conditions and operating condition data vectors within the precursor collection period are eliminated from the historical operating condition sequence. The historical operating condition sequence is supplemented by interpolation and normalized to obtain a time series operating condition training set. The time series operating condition training set within the monitoring window is intercepted to obtain a window operating condition sequence. Use long short-term memory (LSTM) networks to build and train time-series current prediction models and time-series operating condition prediction models; The time series current prediction model is configured with an input layer, a hidden layer, and an output layer. The input layer receives the window current sequence, the hidden layer has two layers of LSTM units, and a dropout layer is added to prevent overfitting. The output layer outputs the time series prediction current sequence of the prediction window through a fully connected layer. During the training process, the root mean square error is used as the loss function, the Adam optimizer is used, the initial learning rate, decay strategy, and training rounds are set, and the early stopping method is used to avoid overfitting. The time series current training set is divided into a training set and a validation set in a ratio of 8:2. The root mean square error (RMSE) of the time series current prediction model on the validation set is calculated. When the RMSE is less than the preset threshold, the time series current prediction model training is considered complete. The prediction window is a period of time starting from the current time and with a constant length. The prediction window slides with the change of time. Input the window current sequence into the trained time series current prediction model to obtain the time series prediction current sequence of the motor in the prediction window; Similarly, the time series working condition prediction model is configured with an input layer, a hidden layer, and an output layer. The input layer receives the window working condition sequence, and the output layer outputs the time series prediction working condition sequence of the prediction window through the fully connected layer. The training is performed using a training set and a validation set that are divided into a ratio of 8:2 using the time series working condition training set. The root mean square error (RMSE) of the time series working condition prediction model on the validation set is calculated. When the RMSE is less than the preset threshold, the training of the time series working condition prediction model is considered complete. Input the window operating condition sequence into the trained time series operating condition prediction model to obtain the time series prediction operating condition sequence of the motor in the prediction window; The basic information of each motor is obtained through the motor nameplate parameters, design data, and mechanical parameters. The loss parameters of each motor are obtained through the historical current series and historical operating condition series. Based on the motor dynamics equations and Kirchhoff's law, combined with the multi-motor coupling effect, a motor group coupling model is established. The input is the operating condition data vector of each motor, and the output is the three-phase stator current vector of each motor: ; The obtained time series prediction operating condition sequence is input into the motor group coupling model to obtain the physical prediction current sequence of each motor; Defining forecast accuracy metrics and , and Respectively represent the root mean square error of the time series current prediction model and the motor group coupling model on the time series current training set, and set the weights for the time series current prediction model and the motor group coupling model and : ; ; The fusion prediction model performs weighted fusion of the prediction results of the time series current prediction model and the motor group coupling model based on the calculated weights, that is, the time series prediction current sequence and the physical prediction current sequence are weightedly fused through the calculated weights to obtain the predicted current sequence of each motor. The predicted current sequence includes the predicted three-phase stator current vector of the motor at each acquisition moment within the prediction window. ; ; in Indicates the current moment, Indicates the Lth sampling moment after the current moment; It should be noted that the purpose of this step is to use historical normal data to train the time series prediction model when no current anomaly is found, and to establish a motor group coupling model based on the basic motor information and loss parameters. The two models are used to predict current and fuse the results. The time series prediction model captures the time series characteristics of motor parameters, and the coupling model considers the interaction between multiple motors. The fusion prediction improves accuracy and predicts current changes in advance, providing data support for coal mine motor operation scheduling and fault prevention. The fusion method dynamically assigns weights based on prediction accuracy, enhancing the reliability and adaptability of the prediction results. S4: Analyze the historical current sequence within the monitoring window and the predicted current sequence, analyze the motor current anomaly and issue an early warning; Based on the window current sequence and the obtained predicted current sequence, the three-phase stator current vector at each acquisition moment in the monitoring window and the prediction window is obtained. and predicted three-phase stator current vector , calculate the effective value of current and predicted current effective value , the formula is: ; ; Set the current abnormal threshold. If there is any current RMS value at any acquisition moment, Or predict the effective value of current If the current is greater than the set abnormal current threshold, it is determined that a current overload abnormality occurs at the acquisition moment. If the current overload abnormality occurs at no less than KL acquisition moments in succession, it is determined that the corresponding motor has a current overload abnormality. Otherwise, calculate the three-phase imbalance of the motor and predicted three-phase imbalance , the formula is: ; ; If there is any acquisition moment that satisfies the three-phase imbalance Or predict three-phase imbalance If it is greater than the national standard critical value, it is determined that a three-phase imbalance abnormality occurs at the acquisition moment. If a three-phase imbalance abnormality occurs at no less than KL acquisition moments in succession, it is determined that a three-phase imbalance abnormality exists in the corresponding motor. If the motor has a current overload abnormality or a three-phase imbalance abnormality, an alarm signal is generated and the motor number and current abnormality type are sent to the administrator terminal. The current abnormality types include current overload abnormality and three-phase imbalance abnormality; It should be noted that the purpose of this step is to determine whether the current is abnormal by calculating indicators such as the effective value of the current and the three-phase imbalance. If abnormal, an alarm will be issued in time to achieve closed-loop monitoring and early warning. The multi-dimensional current abnormality judgment indicators ensure comprehensive abnormality identification. The timely alarm mechanism enables staff to respond quickly, reduce failure losses, and continuously monitor and ensure real-time control of the motor's operating status. The technical solution of an embodiment of the present invention is: setting up a sensor network, collecting the historical current sequence and historical operating condition sequence of each motor, marking abnormal operating conditions and corresponding precursory operating conditions in the historical operating condition sequence of each motor, setting a monitoring window, judging whether there are abnormal operating conditions or precursory operating conditions in the monitoring window, if not, triggering current prediction, and constructing a fusion prediction model, the fusion prediction model includes a time series current prediction model and a time series operating condition prediction model trained based on the historical current sequence and the historical operating condition sequence respectively, and obtaining the basic information of each motor to establish a motor group coupling model, predicting each motor through the fusion prediction model, obtaining a predicted current sequence of each motor, analyzing the historical current sequence and the predicted current sequence in the monitoring window, analyzing the current anomaly of the motor and issuing an early warning.

[0011] The basic principles, main features, and advantages of the present invention are shown and described above. Those skilled in the art should understand that the present invention is not limited to the foregoing embodiments. The foregoing embodiments and descriptions are merely illustrative of the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. A current linkage early warning method for multi-motor coupling effect in coal mines, characterized by: include: A sensor network is set up for the multi-motor linkage system in underground coal mines to collect the historical current sequence and historical operating condition sequence of each motor; Mark abnormal working conditions and corresponding foreboding working conditions in the historical working condition sequence of each motor, set a monitoring window, and determine whether there are abnormal working conditions or foreboding working conditions in the monitoring window. If not, trigger current prediction; If current prediction is triggered, a fusion prediction model is constructed. The fusion prediction model includes a time series current prediction model and a time series operating condition prediction model trained based on historical current sequences and historical operating condition sequences, respectively, as well as a motor group coupling model established by obtaining basic information of each motor. The fusion prediction model is used to predict each motor and obtain a predicted current sequence for each motor. The historical current sequence within the monitoring window is analyzed with the predicted current sequence, the abnormal current of the motor is analyzed and an early warning is issued.

2. The current linkage early warning method for multi-motor coupling effect in coal mines according to claim 1, characterized in that: The historical current sequence and the historical operating condition sequence are obtained as follows: A sensor network is set up for the multi-motor linkage system in underground coal mines. The motor parameters of each motor are collected in real time according to the set data collection frequency. The time when the sensor network collects data in real time according to the data collection frequency is marked as the collection moment. At each collection moment, the three-phase stator current vector and operating condition data vector of each motor are constructed. The period between the moment when the sensor network starts collecting data and the current moment is marked as the historical period. For each motor, the three-phase stator current vector and the operating condition data vector collected during the historical period are integrated according to the time sequence to obtain the historical current sequence and historical operating condition sequence of each motor.

3. The current linkage early warning method for multi-motor coupling effect in coal mines according to claim 2, characterized in that: The method for determining whether an abnormal operating condition or a premonitory operating condition exists is as follows: Set a monitoring window with the current time as the end point and a constant duration, and determine whether there is overlap between the monitoring window and the time period corresponding to any abnormal working condition. If so, determine that there is an abnormal working condition in the monitoring window; If it does not exist, set several omen analysis periods in the monitoring window, all of which end at the current time and have different durations, intercept the historical operating condition sequences of the motor in all omen analysis periods, obtain the omen analysis sequences of the motor, and normalize them respectively to obtain a normalized omen analysis sequence; The normalized precursor sequences of all precursor conditions are obtained. If the DTW distance between any normalized precursor analysis sequence and any normalized precursor sequence within the monitoring window is less than a distance threshold, it is determined that a precursor condition exists within the monitoring window.

4. The current linkage early warning method for multi-motor coupling effect in coal mines according to claim 3, characterized in that: The abnormal operating condition is determined as follows: Obtain and calculate the mean and standard deviation of each parameter of the operating condition data vector in the historical operating condition sequence of each motor within the normal operating condition data range. Determine the dynamic abnormality threshold through data processing. If any parameter of the operating condition data vector at any acquisition moment exceeds the set dynamic abnormality threshold, assign an abnormal moment mark including the motor number to the acquisition moment. For any motor number, if there are abnormal moment marks at no less than KL consecutive collection moments, the corresponding motor is determined to be in an abnormal operating condition, and the collection moments with consecutive abnormal moment marks are defined as an abnormal operating condition of the corresponding motor. Adjacent abnormal operating conditions of the same motor with a time distance less than the preset time threshold are merged into one abnormal operating condition.

5. The current linkage early warning method for multi-motor coupling effect in coal mines according to claim 3, characterized in that: The method for judging the premonition working condition is: Obtain all the precursor sequence groups and the normalized precursor sequences they contain, process the number of normalized precursor sequences in the same precursor sequence group and the total number of abnormal sequences in the corresponding abnormal working conditions of the same type to obtain a stable ratio; If the DTW distance between any two normalized omen sequences that do not belong to the same omen sequence group meets the similarity criterion, the two omen sequence groups are classified into the same similarity set, the number of omen sequence groups in the similarity set where the omen sequence group belongs is obtained, and the data is processed with the total number of all omen sequence groups to obtain the similarity ratio; If the stability ratio is greater than the preset stability standard and the similarity ratio is less than the preset similarity standard, the precursor sequence corresponding to the normalized precursor sequence in the precursor sequence group is determined to be the precursor working condition.

6. The current linkage early warning method for multiple motor coupling effects in coal mines according to claim 5, characterized in that: The method for obtaining the precursor sequence group is as follows: In the motor's historical operating condition sequence, the motor's abnormal sequence is intercepted and normalized respectively to obtain a normalized abnormal sequence and the normalized operating condition data vector. The DTW distance between any two abnormal sequences of the same motor and the corresponding normalized abnormal sequence is calculated. If the distance is less than a distance threshold, the two abnormal sequences are of the same type of abnormal operating condition. In the motor's historical operating condition sequence, the motor's precursor sequences are intercepted and normalized to obtain normalized precursor sequences. Within the same type of abnormal operating condition, the DTW distance between any two abnormal sequences corresponding to the normalized precursor sequences is calculated. If the distance is less than a distance threshold, they are classified into the same precursor sequence group.

7. The current linkage early warning method for multi-motor coupling effect in coal mines according to claim 2, characterized in that: The predicted current sequence is obtained as follows: Obtain the time series current training set, calculate the root mean square error of the time series current prediction model and the motor group coupling model on the time series current training set, and set weights for the time series current prediction model and the motor group coupling model according to the root mean square error. The fusion prediction model performs weighted fusion on the prediction results of the time series current prediction model and the motor group coupling model based on the set weights, that is, the time series prediction current sequence and the physical prediction current sequence are weightedly fused through the calculated weights to obtain the predicted current sequence of each motor.

8. The current linkage early warning method for multiple motor coupling effects in coal mines according to claim 7, characterized in that: The method for obtaining the time series prediction current sequence is as follows: Eliminate all abnormal and foreboding operating condition data vectors and three-phase stator current vectors from the historical operating condition sequence and historical current sequence. Use interpolation to supplement and normalize the historical operating condition sequence and historical current sequence to obtain a time-series operating condition training set and a time-series current training set. Obtain the monitoring window and intercept it to obtain the window operating condition sequence and window current sequence. Using the long short-term memory network, the time series working condition training set and the time series current training set are used to construct and train the time series working condition prediction model and the time series current prediction model respectively. Combining the window working condition sequence and the window current sequence, the time series prediction working condition sequence and the time series prediction current sequence of the motor are obtained.

9. The current linkage early warning method for multiple motor coupling effects in coal mines according to claim 8, characterized in that: The physical prediction current sequence is obtained as follows: The basic information of each motor is obtained, and the loss parameters of each motor are obtained through statistics of historical current sequences and historical operating condition sequences. Based on the motor dynamics equations and Kirchhoff's laws, combined with the multi-motor coupling effect, a motor group coupling model is established with the input as the operating condition data vector of each motor and the output as the three-phase stator current vector of each motor. The time-series predicted operating condition sequence of each motor is input into the motor group coupling model to obtain the physically predicted current sequence of each motor.

10. The current linkage early warning method for multiple motor coupling effects in coal mines according to claim 8, characterized in that: The analysis of the abnormal current of the motor includes: Obtain a window current sequence and a predicted current sequence. The window current sequence includes the three-phase stator current vector, and the predicted current sequence includes the predicted three-phase stator current vector. Calculate the current effective value and the predicted current effective value through data processing. If the current effective value or the predicted current effective value meets the current overload abnormality condition for at least KL consecutive acquisition moments, then determine that the corresponding motor has a current overload abnormality. Otherwise, the three-phase imbalance of the motor and the predicted three-phase imbalance are calculated. If the three-phase imbalance or the predicted three-phase imbalance is greater than the national standard critical value at no less than KL consecutive acquisition moments, it is determined that the corresponding motor has a three-phase imbalance abnormality.

Images