Intelligent mine management method based on Internet of Things

Abstract

The invention relates to the technical field of smart mines, and discloses a smart mine management method based on Internet of Things, which comprises the following steps: step 1, arranging sensors on key parts of mine equipment, the sensors comprising a vibration sensor, a temperature sensor and a current sensor and being used for collecting dynamic data generated in the operation process of the equipment, the dynamic data comprises a vibration signal, a temperature signal and a current signal, the collected dynamic data is transmitted to the edge computing device by using an Internet of Things communication protocol, and the edge computing device carries out unified preprocessing on the dynamic data. A wavelet transform method is introduced into dynamic data preprocessing, multi-scale noise filtering is carried out on vibration signals, temperature signals and current signals, environment interference signals are removed, high-purity processing of dynamic data under complex working conditions is achieved, and the effect of clearly presenting weak fault features in the dynamic signals is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of smart mines, and in particular to a smart mine management method based on the Internet of Things. Background Art

[0002] With the rapid development of the mining industry and the popularization of digital and intelligent technologies, mining companies are facing multiple challenges in how to improve production efficiency, reduce production costs and ensure safe production. The mining operating environment is usually complex and harsh. When the equipment is running, it is easily affected by environmental factors such as dust, high temperature, high humidity, vibration and impact, resulting in reduced operating stability of the equipment. At the same time, mining equipment usually operates continuously and at high intensity. Once a failure occurs, it will lead to interruption of the production process and cause safety accidents. Therefore, real-time monitoring and intelligent management of mining equipment, prediction of potential failures and timely warning have become important issues for improving mining production efficiency and safety.

[0003] In traditional mine management, equipment status monitoring usually relies on periodic manual inspections, single sensor signal analysis, and simple alarm systems, which expose obvious problems when faced with complex working conditions. On the one hand, manual inspections have obvious lags and cannot grasp the equipment operating status in real time. On the other hand, traditional single signal analysis methods cannot cope with the nonlinear relationship of multiple variables in equipment operation, resulting in the accuracy and reliability of status monitoring failing to meet actual needs. In addition, due to the particularity of mining operating conditions, equipment operating data often contain high-frequency noise, environmental interference, and other problems, which further increases the difficulty of equipment status monitoring and fault prediction.

[0004] The development of Internet of Things technology, through the deployment of multi-source sensors, can realize dynamic data collection of equipment operation status, providing a data basis for intelligent management of equipment. However, how to remove noise, extract key features, establish accurate health status models and provide early warning for equipment from the collected multi-source data is still a core problem that cannot be overcome by existing technologies. In response to a series of problems, it is urgent to introduce advanced data processing methods and nonlinear modeling technologies to build an efficient and reliable smart mine management method to improve the accuracy of equipment operation monitoring and fault prediction capabilities.

[0005] The following is a specific analysis of the main problems existing in traditional mine management technology:

[0006] The mining operating environment is complex, and the dynamic data generated during equipment operation is easily interfered by external noise. The interference signal is usually included in the original data collected by the equipment, resulting in a decrease in the authenticity and effectiveness of the data, seriously affecting the accuracy of subsequent equipment health status assessments.

[0007] The operating status of mining equipment is usually affected by multiple variables and is highly nonlinear and dynamic. Traditional linear analysis methods cannot effectively capture complex coupling relationships and cannot comprehensively analyze and predict the operating status and failure trends of equipment.

[0008] Existing equipment monitoring methods are mostly based on simple statistical features and fixed alarm thresholds. For example, an alarm is triggered when the temperature or vibration value of the equipment exceeds a certain set value. These methods can identify significant anomalies in equipment operation, but cannot effectively identify the slow deterioration of the equipment's operating status or precursor characteristics, resulting in a reduction in the early warning capability of equipment failure.

[0009] In actual mine management, traditional fault warning methods usually rely on fixed threshold rules or manual experience, ignoring the dynamic changes of equipment operating status and working environment, and are unable to dynamically adjust warning parameters and adapt to the complex and changeable working conditions of mines.

[0010] Therefore, those skilled in the art provide a smart mine management method based on the Internet of Things to solve the above-mentioned problems. Summary of the invention

[0011] In view of the deficiencies in the prior art, the present invention provides a smart mine management method based on the Internet of Things to solve the problems raised in the above background technology.

[0012] To achieve the above objectives, the present invention is implemented through the following technical solutions: a smart mine management method based on the Internet of Things, comprising:

[0013] Step 1: Arrange sensors on key components of mining equipment. The sensors include vibration sensors, temperature sensors, and current sensors. They are used to collect dynamic data generated during the operation of the equipment. The dynamic data includes vibration signals, temperature signals, and current signals. The collected dynamic data is transmitted to the edge computing device using the Internet of Things communication protocol. The edge computing device performs unified preprocessing on the dynamic data.

[0014] Step 2: During the dynamic data preprocessing process, the vibration signal, temperature signal and current signal are subjected to noise filtering and normalization processing. Noise filtering is used to remove environmental interference signals, and normalization processing is used to standardize the data of different physical quantities to the same numerical range. The normalized data obtained will be used as input for multi-scale feature extraction.

[0015] Step 3: Perform multi-scale feature extraction based on the preprocessed normalized data. By analyzing the components of the dynamic data in time and frequency, extract the characteristic quantities of the equipment operation status, including the frequency distribution characteristics and energy distribution characteristics of the signal in different time scales, and store and associate the characteristic quantities with the original dynamic data.

[0016] Step 4: Combine the feature quantities obtained by multi-scale feature extraction, use high-dimensional dynamic modeling methods to reconstruct the time series data, embed the dynamic data into the high-dimensional state space, form a high-dimensional state vector, and identify the evolution law of the equipment operation state by analyzing the change trend of the state vector, and extract the key characteristics reflecting the health status of the equipment;

[0017] Step 5: Based on the equipment operation status characteristics extracted by high-dimensional dynamic modeling, combined with the operating environment and historical operation data of the mining equipment, establish an equipment health status model, and determine whether the equipment is in an abnormal state according to the equipment operation status model. By judging the equipment status category output by the status model, determine whether the equipment operation status is normal;

[0018] Step 6: Combine the output of the equipment health status model with dynamic data and characteristic quantities to generate a health assessment report for the mining equipment, including the current operating status, abnormal status determination results, and equipment operating trend information. At the same time, the report is transmitted to the remote monitoring center and visualization platform for display and further analysis of the health status of the mining equipment.

[0019] Preferably, the noise filtering in step 2 adopts a wavelet transform method, and the dynamic data is decomposed and reconstructed by an appropriate wavelet basis function to remove high-frequency noise in the dynamic data and obtain a pure signal for normalization processing. The signal reconstruction formula of the wavelet transform is:

[0020]

[0021] Among them, W j，k is the wavelet coefficient, B j，k (t) is the wavelet basis function, j is the decomposition scale, k is the time position, x clean (t) represents the pure signal after denoising by wavelet transform, t represents the time variable, and N represents the number of scales of wavelet decomposition.

[0022] Preferably, the normalization process in step 2 is completed by the following formula:

[0023]

[0024] Among them, x i (t) is the original dynamic data,

[0025] min(x i ) and max(x i ) are the minimum and maximum values ​​of the dynamic data, and the normalized data Normalized to the interval [0, 1].

[0026] Preferably, the multi-scale feature extraction in step 3 includes analyzing the frequency distribution characteristics and energy distribution characteristics of the dynamic data at different time scales, and the energy distribution characteristics are calculated by the following formula:

[0027] E j =∑ k |W f (a j , b k )| 2 ,

[0028] Among them, E j Indicated on scale a j The signal energy under f (a j , b k ) is the dynamic data at scale a j and time shift b k The wavelet transform coefficients are:

[0029] Preferably, the high-dimensional dynamic modeling in step 4 uses a time-delayed coordinate method to reconstruct the time series data in a high-dimensional space, and the high-dimensional state vector is expressed as:

[0030] X(t)=[x(t),x(t+r),x(t+2r),…,x(t+(m-1)r)],

[0031] Where r is the delay time, X(t) represents the embedding form of time series data in high-dimensional space, x(t) represents the state feature value of the device at the current moment, and m represents the number of dimensions in the high-dimensional state space;

[0032] Determined by the autocorrelation function of the time series, the following conditions are met:

[0033]

[0034] Among them, ACF(r) represents the correlation between samples delayed by time r in the quantitative time series, where r is the delay time. Used to determine the decay rate of the correlation of time series.

[0035] Preferably, the high-dimensional state vector in step 4 is used to extract the fractal dimension of the equipment operation state, and the fractal dimension is defined as:

[0036]

[0037] Where N(∈) represents the minimum number of boxes covering the device operating state attractor at scale ∈, ∈ represents the scale used to cover the high-dimensional attractor trajectory, and D f Represents the geometric complexity of quantifying the operating state of a device.

[0038] Preferably, the equipment health status model in step 5 combines the equipment operation status characteristics and historical data, and calculates the health status of the equipment through the state entropy of the dynamic system. The formula of the state entropy is:

[0039]

[0040] Among them, S(t) represents the state entropy of the system at time t, S(t) represents the probability distribution of the equipment operation state variable, and n is the number of distribution intervals of the variable.

[0041] Preferably, the health assessment report in step 6 includes an operating status trend curve and a status entropy change curve, and the health assessment report is generated into a visual chart through a remote monitoring center for mine managers to make subsequent maintenance decisions;

[0042] The generated health assessment report is used to trigger a maintenance warning. If the warning conditions are met, the operation of the relevant equipment is stopped through a control instruction and the maintenance personnel are notified.

[0043] Preferably, the warning conditions include: abnormal equipment operation status and remaining life determination;

[0044] The equipment operation state abnormality judgment formula is: the abnormality of the equipment operation state is determined by the abnormal probability P and the abnormal threshold P output by the health state model. threshold It is relatively certain that the formula is:

[0045] Condition 1 :P>P threshold ,

[0046] Among them, P is the abnormal operation probability output by the equipment health status model, P threshold The threshold for determining abnormal operation;

[0047] The remaining life determination formula is: The calculation of the remaining life R is based on the dynamic evolution rate of the health state model. When the predicted remaining life R is less than the preset threshold R threshold When the warning is triggered, the formula is: Condition 2 : R <R threshold ,

[0048] Among them, R is the remaining life, which means the remaining available time predicted for the equipment under the current operating state;

[0049] R threshold It is the warning threshold of remaining life.

[0050] Preferably, the comprehensive formula for triggering the early warning is:

[0051]

[0052] Among them, Prewarning Trigger = 1, indicating that the warning is triggered;

[0053] Prewarning Trigger=0, indicating that no warning is triggered.

[0054] The present invention provides a smart mine management method based on the Internet of Things. It has the following beneficial effects:

[0055] 1. The present invention introduces the wavelet transform method in dynamic data preprocessing, performs multi-scale noise filtering on vibration signals, temperature signals and current signals, removes environmental interference signals, achieves high-purity processing of dynamic data under complex working conditions, and obtains the effect of clearly presenting weak fault characteristics in dynamic signals.

[0056] 2. The present invention embeds dynamic data into high-dimensional state space through the method of time delay embedding and high-dimensional phase space reconstruction, generates high-dimensional state vectors, analyzes the evolution law of equipment operation status, realizes modeling of nonlinear dynamic behaviors hidden in equipment operation, and obtains the effect of accurate description of complex operation status and early identification of abnormal trends.

[0057] 3. The present invention quantifies the geometric complexity of high-dimensional state vectors by using fractal dimensions, analyzes the evolution trend of the equipment operating status on high-dimensional attractors, and achieves an accurate description of the evolution process of the equipment from a normal state to a fault state, thereby achieving the effect of rapid identification of fault precursor states and visualization of changes in equipment health states.

[0058] 4. The present invention combines the abnormal probability and dynamic calculation of the remaining life in the equipment health status model to set comprehensive warning conditions, realize dynamic adjustment of warning parameters under different equipment and working conditions, and obtain early warning capabilities for faults and significantly reduce false alarm and missed alarm rates. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0060] In order to make the technical personnel in the technical field understand the scheme of the present invention, the technical scheme in the embodiment of the present invention will be clearly and completely described below in combination with the drawings in the embodiment of the present invention. Obviously, the described embodiment is a partial embodiment of the present invention, not a complete embodiment. Based on the embodiment of the present invention, other embodiments obtained by ordinary technicians in the field without creative work should fall within the scope of protection of the present invention.

[0061] The present invention is described in detail below in conjunction with the accompanying drawings:

[0062] Example:

[0063] Please refer to the attached Figure 1 The embodiment of the present invention provides a smart mine management method based on the Internet of Things, including:

[0064] Step 1: Arrange sensors on key components of mining equipment. The sensors include vibration sensors, temperature sensors, and current sensors. They are used to collect dynamic data generated during the operation of the equipment. The dynamic data includes vibration signals, temperature signals, and current signals. The collected dynamic data is transmitted to the edge computing device using the Internet of Things communication protocol. The edge computing device performs unified preprocessing on the dynamic data.

[0065] Step 2: During the dynamic data preprocessing process, the vibration signal, temperature signal and current signal are subjected to noise filtering and normalization processing. Noise filtering is used to remove environmental interference signals, and normalization processing is used to standardize the data of different physical quantities to the same numerical range. The normalized data obtained will be used as input for multi-scale feature extraction.

[0066] Step 3: Perform multi-scale feature extraction based on the preprocessed normalized data. By analyzing the components of the dynamic data in time and frequency, extract the characteristic quantities of the equipment operation status, including the frequency distribution characteristics and energy distribution characteristics of the signal in different time scales, and store and associate the characteristic quantities with the original dynamic data.

[0067] Step 4: Combine the feature quantities obtained by multi-scale feature extraction, use high-dimensional dynamic modeling methods to reconstruct the time series data, embed the dynamic data into the high-dimensional state space, form a high-dimensional state vector, and identify the evolution law of the equipment operation state by analyzing the change trend of the state vector, and extract the key characteristics reflecting the health status of the equipment;

[0068] Step 5: Based on the equipment operation status characteristics extracted by high-dimensional dynamic modeling, combined with the operating environment and historical operation data of the mining equipment, establish an equipment health status model, and determine whether the equipment is in an abnormal state according to the equipment operation status model. By judging the equipment status category output by the status model, determine whether the equipment operation status is normal;

[0069] Step 6: Combine the output of the equipment health status model with dynamic data and characteristic quantities to generate a health assessment report for the mining equipment, including the current operating status, abnormal status determination results, and equipment operating trend information. At the same time, the report is transmitted to the remote monitoring center and visualization platform for display and further analysis of the health status of the mining equipment.

[0070] Benefits of step 1: Realize multi-source collection of dynamic data to ensure the comprehensiveness and accuracy of the equipment operation status. Realize real-time data transmission to provide guarantee for subsequent data processing and analysis. The diversity of dynamic data collection can comprehensively reflect the equipment operation status and improve the ability to capture equipment operation behavior;

[0071] Benefits of step 2: Noise filtering removes environmental interference signals, making the collected dynamic data pure and reducing the impact of data noise on feature extraction and modeling results. Normalization processing standardizes the data of different physical quantities to the same numerical range, solves the problem of inconsistent scales of multivariate data, and improves the computational adaptability of subsequent analysis. By preprocessing dynamic data, the reliability and accuracy of the feature extraction process are ensured, laying the foundation for equipment operation status analysis;

[0072] Benefits of step 3: By analyzing the components of dynamic data in time and frequency, the frequency distribution characteristics and energy distribution characteristics of the signal can be extracted, which can accurately capture the multi-dimensional characteristics of the equipment's operating status. The extracted feature quantities fully reflect the equipment's operating status and enhance the depth of equipment status monitoring. The extracted feature quantities are associated with the original dynamic data and stored to provide basic data for subsequent high-dimensional dynamic modeling;

[0073] Benefits of step 4: By embedding dynamic data into a high-dimensional state space and generating a high-dimensional state vector, the multivariate dynamic behavior of the equipment's operating state can be fully described. High-dimensional dynamic modeling can capture the nonlinear evolution of equipment operation, especially when the equipment gradually evolves to a fault state, it can identify hidden state change trends. The extracted key health status characteristics provide core data support for establishing the equipment health status model, improving the accuracy of the description of the equipment's operating status;

[0074] Benefits of step 5: Combine the equipment operating status characteristics, operating environment and historical operating data extracted by high-dimensional dynamic modeling to establish a highly adaptable health status model to achieve scientific quantification of the equipment health status. The equipment health status model can determine whether the equipment is in an abnormal state, accurately distinguish between normal operating status and potential fault status, and provide data support for early fault identification. The output of the health status model directly supports the early warning system, improving the intelligence and accuracy of equipment operating status monitoring;

[0075] Benefits of step 6: The output of the equipment health status model is combined with dynamic data and feature quantities to generate a health assessment report, which provides the current operating status of the equipment, abnormal status determination, and operating trend information, allowing managers to quickly understand the health status of the equipment. The visual display of the health assessment report is transmitted through the remote monitoring center and platform, making the equipment operating status more intuitive and understandable.

[0076] The noise filtering in step 2 uses the wavelet transform method to decompose and reconstruct the dynamic data through appropriate wavelet basis functions to remove the high-frequency noise in the dynamic data and obtain a pure signal for normalization processing. The signal reconstruction formula of the wavelet transform is:

[0077]

[0078] Among them, W j，k is the wavelet coefficient, B j，k (t) is the wavelet basis function, j is the decomposition scale, k is the time position, x clean (t) represents the pure signal after denoising by wavelet transform, t represents the time variable, and N represents the number of scales of wavelet decomposition.

[0079] By decomposing dynamic data into multi-scale time-frequency components through wavelet transform, high-frequency noise can be effectively separated from the main characteristic signals of dynamic data, and noise can be removed through signal reconstruction. Multi-scale analysis methods are more targeted and flexible than traditional filters, suitable for processing non-stationary signals, making the collected data pure and significantly improving the accuracy of subsequent data processing and analysis.

[0080] Compared with the traditional full-band filtering method, wavelet transform can retain the key characteristic signals in dynamic data on the basis of multi-scale decomposition, while eliminating irrelevant high-frequency noise and interference. Through the denoising process of wavelet transform, the integrity of the signal in the time domain and frequency domain can be ensured, laying a data foundation for subsequent feature extraction.

[0081] The operating environment of mining equipment is complex, and the signals contain nonlinear characteristics and non-stationary components. Wavelet transform has good localized analysis capabilities and can process time and frequency information at the same time, solving the problem that traditional Fourier transform can only process linear stationary signals.

[0082] The normalization process in step 2 is completed by the following formula:

[0083]

[0084] Among them, x i (t) is the original dynamic data,

[0085] min(x i ) and max(x i ) are the minimum and maximum values ​​of the dynamic data, and the normalized data Normalized to the interval [0, 1].

[0086] Normalization processing unifies dynamic data into the interval [0, 1] to solve the impact of different physical quantity scale differences on calculation accuracy, and at the same time, improve the stability and adaptability of data processing and model calculation. Normalization improves the accuracy of subsequent feature extraction and state modeling, and significantly improves the training efficiency and prediction performance of the model, providing important guarantees for the efficient operation of the smart mine management system. Overall, normalization is an indispensable and important link in the dynamic data preprocessing process, laying the foundation for data consistency and reliability of the management process.

[0087] The multi-scale feature extraction in step 3 includes analyzing the frequency distribution characteristics and energy distribution characteristics of dynamic data at different time scales. The energy distribution characteristics are calculated by the following formula:

[0088] E j =∑ k |W f (a j , b k )| 2 ,

[0089] Among them, E j Indicated on scale a j The signal energy under f (a j , b k ) is the dynamic data at scale a j and time shift b k The wavelet transform coefficients are:

[0090] Multi-scale feature extraction can fully reveal the multi-dimensional characteristics of the operating status of mining equipment and effectively extract key features related to faults by analyzing the frequency distribution and energy distribution of dynamic data at different time scales. The energy distribution characteristics quantify the characteristic changes of the signal, making the equipment status description more accurate. At the same time, it adapts to the dynamic change requirements under complex working conditions and provides input data for subsequent high-dimensional dynamic modeling and health status assessment. It is an important part of the smart mine management process.

[0091] The high-dimensional dynamic modeling in step 4 uses the time-delay coordinate method to reconstruct the time series data in high-dimensional space. The high-dimensional state vector is expressed as:

[0092] X(t)=[x(t),x(t+r),x(t+2r),…,x(t+(m-1)r)],

[0093] Where r is the delay time, X(t) represents the embedding form of time series data in high-dimensional space, x(t) represents the state feature value of the device at the current moment, and m represents the number of dimensions in the high-dimensional state space;

[0094] Determined by the autocorrelation function of the time series, the following conditions are met:

[0095]

[0096] Among them, ACF(r) represents the correlation between samples delayed by time r in the quantitative time series, where r is the delay time. Used to determine the decay rate of the correlation of time series.

[0097] High-dimensional dynamic modeling through the time delay coordinate method can accurately capture the complex dynamic characteristics of the equipment's operating status, quantify the multi-variable time series correlation, and dynamically adapt to the needs of different equipment and working conditions. At the same time, the introduction of the autocorrelation function ensures the optimization of the delay time, avoids the introduction of redundant information, improves modeling efficiency and accuracy, and reveals the characteristic evolution process of the equipment's operating status in high-dimensional space, providing important basic data support for health status assessment and fault prediction, which is an indispensable core link in the smart mine management method.

[0098] The high-dimensional state vector in step 4 is used to extract the fractal dimension of the equipment operation state. The fractal dimension is defined as:

[0099]

[0100] Where N(∈) represents the minimum number of boxes covering the device operating state attractor at scale ∈, ∈ represents the scale used to cover the high-dimensional attractor trajectory, and D f Represents the geometric complexity of quantifying the operating state of a device.

[0101] Extracting fractal dimension based on high-dimensional state vector provides a powerful nonlinear quantification tool for equipment operation status. Fractal dimension can accurately reflect the geometric complexity changes of equipment health status, reveal the hidden fault precursor characteristics in high-dimensional dynamic characteristics, and provide the ability to dynamically adapt to different working conditions and equipment characteristics. Through the quantification of fractal dimension, the evolution process of equipment from normal state to fault state can be intuitively reflected, providing support for health status assessment and fault prediction, effectively making up for the shortcomings of traditional linear modeling methods, and is an important link in dealing with nonlinear complex system behaviors in smart mine management methods.

[0102] The equipment health status model in step 5 combines the equipment operation status characteristics and historical data to calculate the equipment health status through the state entropy of the dynamic system. The formula for the state entropy is:

[0103]

[0104] Among them, S(t) represents the state entropy of the system at time t, S(t) represents the probability distribution of the equipment operation state variable, and n is the number of distribution intervals of the variable.

[0105] By calculating the health status of equipment through dynamic system state entropy, the disorder of the equipment operation status can be accurately quantified, the evolution trend of the equipment health status can be dynamically reflected, and the early capture capability of fault precursors can be enhanced. The multivariable adaptability and dynamic adjustment capability of state entropy make it an important tool for equipment health status assessment under complex working conditions. At the same time, combined with the personalized modeling capability and computing efficiency of historical data, it provides efficient and reliable technical support for smart mine management, plays a core role in health status assessment, and provides a key basis for the prediction and early warning of equipment failures.

[0106] The health assessment report in step 6 includes an operating status trend curve and a status entropy change curve, and the health assessment report is generated into a visual chart through the remote monitoring center for mine managers to make subsequent maintenance decisions;

[0107] The generated health assessment report is used to trigger maintenance warnings. If the warning conditions are met, the operation of related equipment is stopped through control instructions and maintenance personnel are notified.

[0108] Early warning conditions include: abnormal equipment operating status and remaining life determination;

[0109] The equipment operation status abnormality judgment formula: The abnormality of the equipment operation status is determined by the abnormal probability P and the abnormal threshold P output by the health status model. threshold It is relatively certain that the formula is:

[0110] Condition 1 :P>P threshold ,

[0111] Among them, P is the abnormal operation probability output by the equipment health status model, P threshold The threshold for determining abnormal operation;

[0112] Remaining life determination formula: The calculation of the remaining life R is based on the dynamic evolution rate of the health state model. When the predicted remaining life R is less than the preset threshold R threshold When the warning is triggered, the formula is: Condition 2 : R <R threshold ,

[0113] Among them, R is the remaining life, which means the remaining available time predicted for the equipment under the current operating state;

[0114] R threshold is the warning threshold of remaining life;

[0115] The comprehensive formula for triggering the early warning is:

[0116]

[0117] Among them, Prewarning Trigger = 1, indicating that the warning is triggered;

[0118] Prewarning Trigger=0, indicating that no warning is triggered.

[0119] The health assessment report and early warning mechanism in step 6 provide comprehensive technical support for equipment operation status monitoring and fault early warning through real-time monitoring, visual display and automated response. The health assessment report intuitively shows the equipment's operating status and trends. The early warning mechanism combines the comprehensive formula for abnormal equipment operation status and remaining life judgment to significantly improve the accuracy and timeliness of the early warning. The mechanism that automatically stops the equipment operation and notifies maintenance personnel after the early warning is triggered shortens the response time and avoids safety risks and economic losses caused by the continued operation of the equipment in an abnormal state. Overall, step 6 enhances the intelligent management capabilities of mining equipment, effectively reduces maintenance costs and downtime, and is an important part of ensuring the safe operation of equipment in the smart mine management method.

[0120] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A smart mine management method based on the Internet of Things, characterized in that: include: Step 1: Arrange sensors on key components of mining equipment. The sensors include vibration sensors, temperature sensors, and current sensors. They are used to collect dynamic data generated during the operation of the equipment. The dynamic data includes vibration signals, temperature signals, and current signals. The collected dynamic data is transmitted to the edge computing device using the Internet of Things communication protocol. The edge computing device performs unified preprocessing on the dynamic data. Step 2: During the dynamic data preprocessing process, the vibration signal, temperature signal and current signal are subjected to noise filtering and normalization processing. Noise filtering is used to remove environmental interference signals, and normalization processing is used to standardize the data of different physical quantities to the same numerical range. The normalized data obtained will be used as input for multi-scale feature extraction. Step 3: Perform multi-scale feature extraction based on the preprocessed normalized data. By analyzing the components of the dynamic data in time and frequency, extract the characteristic quantities of the equipment operation status, including the frequency distribution characteristics and energy distribution characteristics of the signal in different time scales, and store and associate the characteristic quantities with the original dynamic data. Step 4: Combine the feature quantities obtained by multi-scale feature extraction, use high-dimensional dynamic modeling methods to reconstruct the time series data, embed the dynamic data into the high-dimensional state space, form a high-dimensional state vector, and identify the evolution law of the equipment operation state by analyzing the change trend of the state vector, and extract the key characteristics reflecting the health status of the equipment; Step 5: Based on the equipment operation status characteristics extracted by high-dimensional dynamic modeling, combined with the operating environment and historical operation data of the mining equipment, establish an equipment health status model, and determine whether the equipment is in an abnormal state according to the equipment operation status model. By judging the equipment status category output by the status model, determine whether the equipment operation status is normal; Step 6: Combine the output of the equipment health status model with dynamic data and characteristic quantities to generate a health assessment report for the mining equipment, including the current operating status, abnormal status determination results, and equipment operating trend information. At the same time, the report is transmitted to the remote monitoring center and visualization platform for display and further analysis of the health status of the mining equipment.

2. According to the method of intelligent mine management based on the Internet of Things in claim 1, it is characterized in that: The noise filtering in step 2 adopts the wavelet transform method to decompose and reconstruct the dynamic data through appropriate wavelet basis functions to remove high-frequency noise in the dynamic data and obtain a pure signal for normalization processing. The signal reconstruction formula of the wavelet transform is: Among them, W j，k is the wavelet coefficient, B j，k (t) is the wavelet basis function, j is the decomposition scale, k is the time position, x clean (t) represents the pure signal after denoising by wavelet transform, t represents the time variable, and N represents the number of scales of wavelet decomposition.

3. The method for intelligent mine management based on the Internet of Things according to claim 2 is characterized in that: The normalization process in step 2 is completed by the following formula: Among them, x i (t) is the original dynamic data, min(x i ) and max(x i ) are the minimum and maximum values ​​of the dynamic data, and the normalized data Normalized to the interval [0, 1].

4. The method for intelligent mine management based on the Internet of Things according to claim 1 is characterized in that: The multi-scale feature extraction in step 3 includes analyzing the frequency distribution characteristics and energy distribution characteristics of dynamic data at different time scales, and the energy distribution characteristics are calculated by the following formula: HAVE BEEN j =∑ k |W f (a j ,b k )| 2 , Among them, E j Indicated on scale a j The signal energy under f (a j , b k ) is the dynamic data at scale a j and time shift b k The wavelet transform coefficients are:

5. The method for intelligent mine management based on the Internet of Things according to claim 4 is characterized in that: The high-dimensional dynamic modeling in step 4 uses the time-delay coordinate method to reconstruct the time series data in high-dimensional space, and the high-dimensional state vector is expressed as: X(t)=[x(t),x(t+r),x(t+2r),…,x(t+(m-1)r)], Where r is the delay time, X(t) represents the embedding form of time series data in high-dimensional space, x(t) represents the state feature value of the device at the current moment, and m represents the number of dimensions in the high-dimensional state space; Determined by the autocorrelation function of the time series, the following conditions are met: Among them, AcF(r) represents the correlation between samples delayed by time r in the quantized time series, where r is the delay time. Used to determine the decay rate of the correlation of time series.

6. The method for intelligent mine management based on the Internet of Things according to claim 5 is characterized in that: The high-dimensional state vector in step 4 is used to extract the fractal dimension of the equipment operation state. The fractal dimension is defined as: Where N(∈) represents the minimum number of boxes covering the device operating state attractor at scale ∈, ∈ represents the scale used to cover the high-dimensional attractor trajectory, and D f Represents the geometric complexity of quantifying the operating state of a device.

7. The method for intelligent mine management based on the Internet of Things according to claim 1, characterized in that: The equipment health status model in step 5 combines the equipment operation status characteristics and historical data to calculate the health status of the equipment through the state entropy of the dynamic system. The formula of the state entropy is: Among them, S(t) represents the state entropy of the system at time t, S(t) represents the probability distribution of the equipment operation state variable, and n is the number of distribution intervals of the variable.

8. The method for intelligent mine management based on the Internet of Things according to claim 1, characterized in that: The health assessment report in step 6 includes an operating status trend curve and a status entropy change curve, and the health assessment report is generated into a visual chart through a remote monitoring center for mine managers to make subsequent maintenance decisions; The generated health assessment report is used to trigger a maintenance warning. If the warning conditions are met, the operation of the relevant equipment is stopped through a control instruction and the maintenance personnel are notified.

9. The method for intelligent mine management based on the Internet of Things according to claim 8, characterized in that: The warning conditions include: abnormal equipment operation status and remaining life determination; The equipment operation state abnormality judgment formula is: the abnormality of the equipment operation state is determined by the abnormal probability P and the abnormal threshold P output by the health state model. threshold It is relatively certain that the formula is: Condition1: P>P threshold , Among them, P is the abnormal operation probability output by the equipment health status model, P threshold The threshold for determining abnormal operation; The remaining life determination formula is: The calculation of the remaining life R is based on the dynamic evolution rate of the health state model. When the predicted remaining life R is less than the preset threshold R threshold When the warning is triggered, the formula is: Condition2: R <R threshold , Among them, R is the remaining life, which means the remaining available time predicted for the equipment under the current operating state; R threshold It is the warning threshold of remaining life.

10. The method for intelligent mine management based on the Internet of Things according to claim 9, characterized in that: The comprehensive formula for triggering the early warning is: Among them, Prewarning Trigger = 1, indicating that the warning is triggered; Prewarning Trigger=0, indicating that no warning is triggered.

Images