Intelligent monitoring system and method for coal unloader

Abstract

The invention relates to the technical field of coal unloaders, in particular to an intelligent monitoring system and method for a coal unloader, and the system comprises an operation data collection module, a priority scheduling module, an abnormal state detection module, an abnormal state response module and a fault prediction module. According to the method, potential equipment problems are recognized in advance by analyzing the fluctuation range and the sudden change point in the time sequence, the early warning function is achieved, key information can be rapidly processed by dynamically adjusting the transmission priority of data, and therefore the reaction time is shortened, the response efficiency is improved, and the method is suitable for popularization and application. Accurate identification and quick response of the abnormal state effectively perform fault assessment by calculating offset and duration so as to accurately process the abnormal condition, through the data analysis and priority management strategy, not only is the operation flow of the coal unloading machine optimized, but also the fault rate of the coal unloading machine is greatly reduced through predictive maintenance, and the working efficiency of the coal unloading machine is improved. And continuity and safety of operation are ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of coal unloaders, and in particular to an intelligent monitoring system and method for a coal unloader. Background Art

[0002] A coal unloader is a heavy mechanical equipment used for unloading, processing, and transporting coal. The equipment is usually used in places such as power plants, ports, and coal mines. Its main function is to efficiently and quickly unload coal from transportation vehicles such as trucks, trains, or ships. The design of the coal unloader varies according to different working environments and requirements, including various structural forms such as rotary, bridge, and gantry types. The machine requires a high degree of precise control and efficiency during operation to reduce energy consumption and operation time, while ensuring operation safety.

[0003] Among them, the intelligent monitoring system of the coal unloader realizes remote monitoring and operation management of the coal unloader. The system integrates sensor data and an operation interface on a cloud platform, allowing operators to monitor the working status, performance parameters, and potential faults of the coal unloader anywhere with a network connection. Its main purpose is to improve operation efficiency, reduce maintenance costs, optimize the coal processing process, and enhance operation safety. Through remote monitoring, not only can equipment problems be responded to in real time, but also data analysis and predictive maintenance can be carried out, thereby preventing faults from occurring and ensuring continuous and stable operation efficiency.

[0004] The prior art lacks effective remote fault diagnosis and predictive maintenance, restricting the ability of rapid data processing and real-time fault response, resulting in only passive handling after problems occur. The traditional technology fails to realize dynamic adjustment of data transmission priorities, leading to the inability to process key information of the equipment in a timely manner, which is likely to cause equipment damage and operation interruption in case of emergencies. In addition, the lack of efficient data priority management and real-time performance monitoring results in insufficient fault prevention measures, unable to effectively predict and prevent faults from occurring, increasing maintenance costs and reducing operation efficiency, and having a negative impact on the maintenance and stable operation of the coal unloader. Summary of the Invention

[0005] The purpose of the present invention is to solve the deficiencies existing in the prior art, and to propose an intelligent monitoring system and method for a coal unloader.

[0006] To achieve the above purpose, the present invention adopts the following technical solution: An intelligent monitoring system for a coal unloader, the system includes:

[0007] An operation data acquisition module, based on a cloud platform, monitors the operation data of the coal unloader, analyzes the fluctuation range of data within a time series, eliminates timestamp fault data, and obtains a status monitoring data set;

[0008] The priority scheduling module detects changes in the drum torque and the temperature of the bucket wheel drive motor based on the state monitoring data set, marks the data with excessive rate changes, adjusts its transmission priority in the data stream, and obtains a priority-adjusted data stream;

[0009] The abnormal state detection module calculates the offset of each parameter based on the priority-adjusted data stream, screens the parameters with critical offsets, and evaluates the abnormal state in combination with the duration of the abnormality to obtain an abnormal score index for the coal unloader;

[0010] The abnormal state response module compares the change situation at the time of the abnormality based on the abnormal score index of the coal unloader, determines the current working load level according to the load change, screens the response parameters that meet the current load level, and adjusts the operation mode of the coal unloader to obtain a response adjustment parameter set;

[0011] The fault prediction module analyzes the parameter change rate before and after the occurrence of the abnormal state based on the response adjustment parameter set, screens the parameters with critical change rates, and predicts the change trend in the future time period to obtain a fault trend prediction result.

[0012] The improvements of the present invention are that the state monitoring data set includes torque data, temperature data, tension force data, and pressure data, the priority-adjusted data stream includes change rate marking information, data fluctuation screening results, and priority adjustment results, the abnormal score index of the coal unloader includes offset calculation results, error value reconstruction information, abnormal scores, and continuous duration analysis results, and the response adjustment parameter set includes state level analysis results, load level determination results, and operation mode adjustment results.

[0013] The improvements of the present invention are that the operation data acquisition module includes:

[0014] The data parsing sub-module collects the drum torque, the temperature of the bucket wheel drive motor, the conveyor belt tension force, and the hydraulic system pressure data based on the cloud platform, parses the single-point information in the data stream, extracts the measurement values at multiple time points, and constructs a time series data set;

[0015] The time series fluctuation analysis sub-module analyzes the fluctuation range of each parameter on the time axis based on the time series data set, calculates the numerical difference between consecutive time points, detects abnormal mutation points, and uses the formula:

[0016]

[0017] Calculate the mutation point fluctuation index V t , screen the points with data fluctuation mutations to obtain data fluctuation mutation points, where X t represents the measurement value at time point t, and X t-1 represents the measurement value at time point t-1, represents the mean of the time series data, and N represents the total number of data points in the time series;

[0018] Based on the data fluctuation mutation points, the status monitoring data screening sub-module eliminates the timestamp fault data, screens the continuous key data intervals, and organizes the screened data to obtain the status monitoring data set.

[0019] The improvement of the present invention is that the priority scheduling module includes:

[0020] The status monitoring sub-module, based on the status monitoring data set, detects the drum torque and the temperature data of the bucket wheel drive motor, calculates the increment values at multiple time points within the time series, and obtains the time series increment data;

[0021] The rate analysis sub-module, based on the time series increment data, calculates the change rate within the time series, using the formula:

[0022]

[0023] By comparing with the normal fluctuation range, single-item data with a rate change exceeding the range is screened to obtain the over-limit rate data, where RS t represents the rate change value, TS i represents the value of the bucket wheel drive motor temperature at time i, MS i represents the value of the drum torque at time i, n S represents the number of sample points within the time series, TS i-1 represents the value of the bucket wheel drive motor temperature at time i - 1, MS i-1 represents the value of the drum torque at time i - 1;

[0024] The priority adjustment sub-module calls the over-limit rate data, conducts a trend analysis of the changes, determines whether there is a continuous increase or abnormal fluctuation situation, and adjusts its transmission priority in the data stream to obtain the priority-adjusted data stream.

[0025] The improvement of the present invention is that the abnormal state detection module includes:

[0026] The parameter offset calculation sub-module, based on the priority-adjusted data stream, collects the real-time data of the conveyor belt tension, the hydraulic system pressure, and the temperature of the bucket wheel drive motor, and calculates the offset amplitude relative to the historical reference value, using the formula:

[0027]

[0028] to obtain the parameter offset amplitude, where ΔZP r represents the offset amplitude of the r-th parameter, ZX r represents the measured value of the r-th parameter, ZX b,rRepresents the historical reference value of the r-th parameter;

[0029] The key parameter screening sub-module screens the parameters with critical offset degrees based on the parameter offset amplitude, calculates their reconstruction error values, determines whether they belong to the abnormal range, and obtains the key offset parameters;

[0030] The abnormal score calculation sub-module calculates the abnormal score based on the key offset parameters, and evaluates the abnormal state in combination with the abnormal duration, and obtains the abnormal score index of the coal unloader.

[0031] The present invention is improved in that the abnormal state response module includes:

[0032] The abnormal level determination sub-module determines the abnormal state level based on the abnormal score index of the coal unloader, compares the torque fluctuation amplitude and the tension force change trend at the abnormal level, and screens the corresponding torque change range and tension force adjustment amplitude to obtain the abnormal state determination result;

[0033] The load change analysis sub-module calculates the load change rate during the operation of the coal unloader based on the abnormal state determination result and the data of the drum torque and the conveyor belt tension force when the abnormality occurs, using the formula:

[0034]

[0035] Screen the load change interval that meets the abnormal level to obtain the current working load level, where LD represents the load change rate, TD m Represents the current drum torque, TD b Represents the reference drum torque, FD s Represents the current conveyor belt tension force, FD b Represents the reference conveyor belt tension force;

[0036] The adjustment parameter screening sub-module compares the current working load level with the adjustment requirements of the abnormal state determination result, screens the response operation parameters of the coal unloader that meet the current state, and adjusts the operation mode of the coal unloader to obtain the response adjustment parameter set.

[0037] The present invention is improved in that the fault prediction module includes:

[0038] The change rate analysis sub-module obtains the data sequences of the hydraulic system pressure, the temperature of the bucket wheel drive motor, and the conveyor belt tension force based on the response adjustment parameter set, calculates the time change rate of the data at adjacent time points, and screens the abnormal fluctuation points to obtain the parameter change rate after screening;

[0039] The key factor screening sub-module analyzes the change trends of each parameter before and after the occurrence of the abnormal state based on the change rate of the screened parameters, calculates the fluctuation amplitude in the differential state, compares it with the rate difference, screens the parameters with critical fluctuation ranges, and forms a set of fault impact factors;

[0040] The fault trend prediction sub-module calculates the fluctuation range of the impact factors in the historical data based on the set of fault impact factors, using the formula:

[0041]

[0042] Predicts the change amount of the fault trend in the future time period to obtain the fault trend prediction result, where TB f represents the fault trend prediction value, WB j is the weight of the jth impact factor, VB j represents the change rate of the jth impact factor, RB j represents the maximum value of the jth impact factor in the historical data, MB j represents the minimum value of the jth impact factor in the historical data, n B is the number of impact factors.

[0043] An intelligent monitoring method for a coal unloader, comprising the following steps:

[0044] S1: Based on the cloud platform, monitor the operation data of the coal unloader, parse the single-point information in the data stream, analyze the fluctuation range of the data within the time series, eliminate the timestamp fault data, and obtain the status monitoring data set;

[0045] S2: Based on the status monitoring data set, detect the changes in the drum torque and the temperature of the bucket wheel drive motor, calculate the change rate within the time series, mark the data with an over-limit rate change, and adjust its transmission priority in the data stream to obtain the priority-adjusted data stream;

[0046] S3: Based on the priority-adjusted data stream, calculate the offset of each parameter within the normal operation range, screen the parameters with critical offset degrees, and evaluate the abnormal state in combination with the abnormal occurrence duration to obtain the abnormal score index of the coal unloader;

[0047] S4: Based on the abnormal score index of the coal unloader, compare the change situation at the time of the abnormality, determine the current working load level according to the load change, screen the response parameters that meet the current load level, and adjust the operation mode of the coal unloader to obtain the response adjustment parameter set;

[0048] S5: Based on the response adjustment parameter set, analyze the parameter change rate before and after the occurrence of the abnormal state, screen the parameters with key change rates, analyze the fluctuation range of the influencing factors in the historical data, predict the change trend in the future time period, and obtain the fault trend prediction result.

[0049] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0050] In the present invention, by analyzing the fluctuation range and mutation points in the time series, potential equipment problems are identified in advance, realizing the early warning function. By dynamically adjusting the transmission priority of data, key information can be processed quickly, thereby reducing the response time and improving the response efficiency. The accurate identification and rapid response of the abnormal state effectively evaluate the fault by calculating the offset and duration, so as to accurately handle the abnormal situation. Through this data analysis and priority management strategy, not only the operation process of the coal unloader is optimized, but also the failure rate of the coal unloader is greatly reduced through predictive maintenance, ensuring the continuity and safety of the operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 is the system flow chart of the present invention;

[0052] Figure 2 is the flow chart of the operation data acquisition module in the present invention;

[0053] Figure 3 is the flow chart of the priority scheduling module in the present invention;

[0054] Figure 4 is the flow chart of the abnormal state detection module in the present invention;

[0055] Figure 5 is the flow chart of the abnormal state response module in the present invention;

[0056] Figure 6 is the flow chart of the fault prediction module in the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0057] In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0058] Embodiment

[0059] Please refer to Figure 1 , the present invention provides a technical solution: an intelligent monitoring system for a coal unloader includes:

[0060] The operation data acquisition module is based on the cloud platform, monitors the operation data of the coal unloader, collects data such as drum torque, temperature of the bucket wheel drive motor, conveyor belt tension, and hydraulic system pressure, analyzes the single-point information in the data stream, analyzes the fluctuation range of the data within the time series, filters out the points where the data fluctuates suddenly, and eliminates the data with timestamp breaks to obtain the status monitoring data set;

[0061] The priority scheduling module is based on the status monitoring data set, detects the changes in the drum torque and the temperature of the bucket wheel drive motor, calculates the change rate within the time series, marks the single-item data whose rate of change exceeds the normal fluctuation range, analyzes the change trend of the marked data, and adjusts its transmission priority in the data stream to obtain the priority-adjusted data stream;

[0062] The abnormal state detection module is based on the priority-adjusted data stream, collects each parameter of the conveyor belt tension, hydraulic system pressure, and temperature of the bucket wheel drive motor, calculates the offset of each parameter within the normal operation range, filters out the parameters with critical offset degrees, calculates the reconstruction error value, and calculates the abnormal score based on the error value. Combining the duration of the abnormality, it evaluates the abnormal state to obtain the abnormal score index of the coal unloader;

[0063] The abnormal state response module is based on the abnormal score index of the coal unloader, analyzes the level of the abnormal state, compares the changes in the drum torque and conveyor belt tension when the abnormality occurs, determines the current working load level of the coal unloader based on the load change, filters out the response parameters that match the current load level, and adjusts the operation mode of the coal unloader to obtain the response adjustment parameter set;

[0064] The fault prediction module is based on the response adjustment parameter set, analyzes the change rates of the hydraulic system pressure, temperature of the bucket wheel drive motor, and conveyor belt tension before and after the occurrence of the abnormal state, filters out the parameters with critical change rates as the fault impact factors, analyzes the fluctuation range of the impact factors in the historical data, and predicts the change trend in the future time period to obtain the fault trend prediction result.

[0065] The status monitoring data set includes torque data, temperature data, tension data, and pressure data. The priority-adjusted data stream includes change rate marking information, data fluctuation screening results, and priority adjustment results. The abnormal score index of the coal unloader includes offset calculation results, error value reconstruction information, abnormal scores, and duration analysis results. The response adjustment parameter set includes status level analysis results, load level determination results, and operation mode adjustment results.

[0066] Please refer to Figure 2 , the operation data acquisition module includes:

[0067] Based on the cloud platform, the data parsing sub-module collects data on the torque of the winding drum, the temperature of the bucket wheel drive motor, the tension of the conveyor belt, and the pressure of the hydraulic system, analyzes the single-point information in the data stream, extracts the measured values at multiple time points, and constructs a time series data set;

[0068] When the coal unloader is in operation, sensors are installed at key positions, such as the winding drum bearing, the surface of the bucket wheel motor, the conveyor belt tensioning device, and the hydraulic pipeline. When collecting data, the sensors obtain the instantaneous values of various physical quantities through measuring elements such as current, temperature, and pressure. The temperature sensor of the bucket wheel drive motor uses a thermocouple to monitor the change of the winding temperature in real time to ensure that the sampling frequency meets the requirements of time series analysis. The data of each sensor is converted from an analog signal to a digital signal and received by the data collection function of the cloud platform. With a sampling frequency of 100Hz, 100 data points are collected per second, which guarantees the integrity of the time series. Subsequently, the single-point information in the data stream is analyzed. Specifically, the original data recorded by the data collection end usually has timestamps. By identifying the timestamps and arranging the data, the measured values of each parameter at multiple time points are extracted. For example, if the temperature of the bucket wheel drive motor at a certain moment is 75.3°C and the data in the next second is 75.5°C, they are stored in chronological order. Finally, by arranging all the measured data points along the time axis, a time series data set is formed. For example, if the torque data of the winding drum within one minute is (1500 Nm, 1520 Nm, 1515 Nm, 1530 Nm, 1525 Nm), then this data sequence can be used for subsequent fluctuation analysis to generate a time series data set.

[0069] Based on the time series data set, the time series fluctuation analysis sub-module analyzes the fluctuation range of each parameter on the time axis, calculates the numerical difference between consecutive time points, detects abnormal mutation points, and uses the formula:

[0070]

[0071] Calculate the mutation point fluctuation index V t , which represents the intensity of data fluctuation at time point t, screens the points where data fluctuation mutates, and obtains the data fluctuation mutation points. Among them, X t represents the measured value at time point t, such as the torque of the winding drum (Nm), the temperature of the bucket wheel drive motor (°C), the tension of the conveyor belt (N), and X t-1 represents the measured value at time point t - 1, represents the mean value of the time series data, and the calculation method is N represents the total number of data points in the time series;

[0072] A time series of a certain measurement data:

[0073] X = [1500, 1520, 1515, 1530, 1525] (unit: Nm);

[0074] Calculate the mean value:

[0075]

[0076] Calculate the sum of squared deviations of each point from the mean value:

[0077]

[0078] Calculate the fluctuation index of a certain mutation point. If X t-1 = 1515, X t = 1530:

[0079]

[0080] Set the mutation point threshold to 0.6. Then, since the mutation index 0.652 exceeds the threshold, it is determined that this data point is a mutation point. In the field of industrial monitoring, such as equipment vibration, torque fluctuation, and temperature change, the identification of mutation points usually uses the normalized fluctuation coefficient for comparison. For industrial equipment operating stably, setting the fluctuation index threshold between 0.5 - 0.7 can effectively balance the false detection rate and the missed detection rate.

[0081] The status monitoring data screening sub-module, based on the data fluctuation mutation points, eliminates the timestamp fault data, screens the continuous key data intervals, and organizes the screened data to obtain the status monitoring data set;

[0082] First, confirm the time continuity of the data. Using the timestamp difference method, calculate the time interval between adjacent data points. For example, in a certain time series, the timestamps are t1 = 0.01s, t2 = 0.02s, t3 = 0.03s, t4 = 0.07s respectively. Then, if the interval between t4 and t3 is more than 3 times the mean value of the previous and subsequent data intervals, it is determined that the t4 point is a fault, and the data corresponding to this time point needs to be eliminated. Subsequently, screen the continuous key data intervals. The determination criterion can be based on the proportion of the data volume within the time period. For example, if the data integrity requirement is set to more than 95%, that is, within a 5-minute time window, at least 28,500 valid data points need to be collected (5min × 60s × 100Hz = 30,000). If the number of valid data is less than 28,500, then the data for this time period is discarded. Finally, organize the screened data, store the parameter data uniformly, format the data structure, and obtain the status monitoring data set.

[0083] Please refer to Figure 3 , the priority scheduling module includes:

[0084] The status monitoring sub-module, based on the status monitoring data set, detects the drum torque and the temperature data of the bucket wheel drive motor, calculates the increment values at multiple time points within the time series, and obtains the time series increment data;

[0085] The drum torque can be measured by a torque sensor installed on the drum shaft. This sensor can sense the torsional deformation of the shaft and convert it into an electrical signal. For example, if the voltage signal output by the torque sensor is 2.5V, the current drum torque can be calculated as 500 N·m by combining the sensor calibration coefficient. The temperature data of the bucket wheel drive motor can be obtained through a thermocouple or a PT100 temperature sensor. For example, if the resistance value detected by the PT100 sensor is 138Ω, the temperature is found to be 50°C according to its temperature-resistance correspondence table. The acquired data is stored according to the time stamp. For each data point, the incremental value between two adjacent time points is calculated. For example, if the drum torque detected at time t1 is 500 N·m and the drum torque detected at time t2 is 520 N·m, the incremental value is calculated as 520 N·m - 500 N·m = 20 N·m. For the temperature of the bucket wheel drive motor, the incremental value at adjacent time points is calculated in the same way. All the calculated incremental values are arranged to form time series incremental data.

[0086] Based on the time series incremental data, the rate analysis sub-module calculates the rate of change within the time series using the formula:

[0087]

[0088] By comparing with the normal fluctuation range, single-item data with a rate of change exceeding the range is screened to obtain out-of-limit rate data, where RS t represents the rate of change value, TS i represents the value of the bucket wheel drive motor temperature at time i, MS i represents the value of the drum torque at time i, n S represents the number of sample points within the time series, TS i-1 represents the value of the bucket wheel drive motor temperature at time i - 1, MS i-1 represents the value of the drum torque at time i - 1;

[0089] Within a certain time window N S = 3, the data collected is as follows: the bucket wheel drive motor temperature data is 55.2, 55.5, and 55.0, and the drum torque data is 1250, 1255, and 1248. Calculate the average absolute value of the temperature increment:

[0090]

[0091] Calculate the square root of the torque increment:

[0092]

[0093] Calculate the rate of change value:

[0094] RS t = RST +RS M =0.2667+8.6023=8.869;

[0095] The result shows that the rate change value in the current time window is 8.869. If it exceeds the set threshold (for example, 1.5), the data in the time window will be marked as excessive rate data and enter the priority adjustment module for data stream transmission optimization.

[0096] The priority adjustment submodule calls the over-limit rate data, performs a change trend analysis, determines whether there is a continuous growth or abnormal fluctuation, and adjusts its transmission priority in the data stream to obtain a priority-adjusted data stream;

[0097] Perform trend analysis on the marked data points to analyze whether there are abnormal conditions such as continuous growth or drastic fluctuations within a certain period of time. For example, in the past 10 minutes, if the change rate of the data point is continuously higher than 45 and increases for three consecutive times, for example, the change value is [46, 48, 50], then it is determined that the data point has a continuous growth trend. If a data point decreases by more than a set threshold compared with the previous time point, for example, the change value is [50, 35], and the decrease exceeds 15, then it is determined that the data point has an abnormal fluctuation trend. According to the trend analysis results, adjust the transmission priority of the data stream. For example, if the proportion of data points exceeding the limit rate exceeds 50%, adjust the transmission priority of the data stream to the highest level to obtain a priority-adjusted data stream.

[0098] See also Figure 4 , the abnormal state detection module includes:

[0099] The parameter offset calculation submodule adjusts the data flow based on the priority, collects the real-time data of the conveyor belt tension, hydraulic system pressure and bucket wheel drive motor temperature, and calculates the offset amplitude relative to the historical baseline value using the formula:

[0100]

[0101] Get the parameter offset amplitude, where ΔZP r Represents the offset amplitude of the rth parameter, ZX r Represents the measured value of the rth parameter, that is, the actual data obtained under the current operating state, ZX b,r represents the historical benchmark value of the rth parameter. The subscript b represents the benchmark, which is used to refer to the reference value of the parameter under normal operating conditions, that is, the reference value under normal operating conditions. It is to calculate the deviation ratio of the current measurement value relative to the historical reference value, that is, the relative rate of change;

[0102] Obtain real-time data of the conveyor belt tension, hydraulic system pressure, and bucket wheel drive motor temperature. The acquisition methods for each parameter are different. For example, the conveyor belt tension can be obtained through a tension sensor installed on the tensioning device, and its unit is Newton (N). The hydraulic system pressure can be collected through a pressure sensor, and its unit is megapascal (MPa). The temperature of the bucket wheel drive motor can be measured through a thermocouple sensor and is expressed in degrees Celsius (°C). The acquired data is calculated after preliminary filtering, calculating the deviation amplitude relative to the historical reference value. The historical reference value refers to the parameter value under normal operation of the equipment and can be calculated through the mean or median of long-term operation data. For example, the reference value of the conveyor belt tension of a certain equipment under normal conditions is 5000 N, the hydraulic system pressure is 10 MPa, and the temperature of the bucket wheel drive motor is 75 °C. After obtaining the measurement data under the current operating state, compare it with the historical reference value to calculate its deviation amplitude. If the current conveyor belt tension is 5300 N, the hydraulic system pressure is 11 MPa, and the temperature of the bucket wheel drive motor is 78 °C, the reference value of the conveyor belt tension is 5000 N, the reference value of the hydraulic system pressure is 10 MPa, and the reference value of the temperature of the bucket wheel drive motor is 75 °C, calculate the deviation amplitude of the conveyor belt tension:

[0103]

[0104] Calculate the deviation amplitude of the hydraulic system pressure:

[0105]

[0106] Calculate the deviation amplitude of the temperature of the bucket wheel drive motor:

[0107]

[0108] The result shows that under the current operating state, compared with the normal operating state of the equipment, the change amplitude of the hydraulic system pressure is the largest (10%), followed by the conveyor belt tension (6%), and the deviation amplitude of the temperature of the bucket wheel drive motor is the smallest (4%), indicating that the hydraulic system is greatly affected. This data can be used for subsequent anomaly screening and anomaly score calculation.

[0109] The key parameter screening sub-module screens the parameters with key deviation degrees based on the parameter deviation amplitude, calculates their reconstruction error values, determines whether they are within the abnormal range, and obtains the key deviation parameters;

[0110] Compare all the offset amplitudes, screen out the parameter with the largest offset amplitude. The calculated offset amplitude of the conveyor belt tension is 6%, the offset amplitude of the hydraulic system pressure is 10%, and the offset amplitude of the temperature of the bucket wheel drive motor is 4%. The parameter with the largest offset amplitude is the hydraulic system pressure. Subsequently, calculate the reconstruction error value for this parameter. The calculation of the reconstruction error value can be based on the sum of squared errors, that is, calculate the sum of squared errors through the deviation between the measured data and the reconstructed data. When the reconstructed value of the hydraulic system pressure is 10.5 MPa and the measured value is 11, the error calculation is ES=(11 - 10.5) 2 = 0.25. This result shows that the change in the hydraulic system pressure not only has the highest offset amplitude (10%), but also has a relatively large error with the reconstructed value (0.25), indicating a relatively high degree of abnormality of this parameter. Compare the reconstruction error value with the set threshold. If it exceeds the threshold, judge whether it belongs to the abnormal range to obtain the key offset parameter.

[0111] The abnormal score calculation sub-module calculates the abnormal score based on the key offset parameter, and evaluates the abnormal state in combination with the abnormal duration to obtain the abnormal score index of the coal unloader;

[0112] Call the offset amplitude of 10% of the hydraulic system pressure and the reconstruction error value of 0.25 to calculate the abnormal score. The abnormal score can be calculated using the normalized score, and calculate the comprehensive abnormal score in combination with the abnormal duration T (minutes). For example, if the abnormal duration is 15 minutes, then calculate the abnormal score as This result shows that during this detection process, the abnormal score of the hydraulic system pressure reaches 5.25. If the abnormal score threshold is 5, it means that the abnormal state of the equipment has reached the warning standard, and it can be used to judge whether abnormal handling measures need to be taken in the future.

[0113] Please refer to Figure 5 , the abnormal state response module includes:

[0114] The abnormal level determination sub-module determines the abnormal state level based on the abnormal score index of the coal unloader, compares the torque fluctuation amplitude and the tension change trend under the abnormal level, and screens the corresponding torque change range and tension adjustment amplitude to obtain the abnormal state determination result;

[0115] Parse multiple monitoring data during the operation of the coal unloader to obtain real-time scoring values. The calculation of the scoring values is based on the operating parameters of the coal unloader, including but not limited to drum torque, conveyor belt tension, vibration frequency, etc. The higher the scoring value, the more serious the degree of abnormality. The specific scoring criteria can be set through historical operation data. For example, the scoring range of a certain type of coal unloader is set from 0 to 100, where 0 - 30 represents normal, 30 - 60 represents mild abnormality, 60 - 90 represents moderate abnormality, and above 90 is severe abnormality. By comparing the current scoring value with the scoring range, determine the level of the abnormal state. Suppose the current scoring value is 72, then it can be determined that the coal unloader is in a moderate abnormal state. After determining the level of the abnormal state, it is necessary to further analyze the torque fluctuation amplitude and the change trend of the conveyor belt tension under the abnormal state. The fluctuation of the drum torque can be reflected by calculating its standard deviation. For example, in the normal state, the standard deviation of the drum torque is 5 N·m, while in the abnormal state, this value rises to 12 N·m. The change trend of the conveyor belt tension can be obtained by calculating its change rate. The calculation method is (the current value of the tension - the reference value of the tension) divided by the reference value of the tension. For example, the reference conveyor belt tension is 1000 N, and the current conveyor belt tension is 1200 N, then its change rate is (1200 - 1000) / 1000 = 0.2, that is, the conveyor belt tension increases by 20%. According to the fluctuation amplitude and the change rate interval corresponding to the abnormal state level, the torque change range and the tension adjustment amplitude that meet the current state can be screened out. For example, in the moderate abnormal state, the drum torque fluctuation range is set from 10 to 15 N·m, and the tension adjustment amplitude is set from 15% to 25%, and the abnormal state determination result is obtained.

[0116] Based on the abnormal state determination result, the load change analysis sub-module calculates the load change rate during the operation of the coal unloader according to the drum torque and conveyor belt tension data at the time of the abnormality, using the formula:

[0117]

[0118] Screen the load change interval that meets the abnormal level to obtain the current working load level. Among them, LD represents the load change rate, which is used to measure the load fluctuation degree during the operation of the coal unloader, TD m represents the current drum torque, which is the drum torque value of the coal unloader in the current state, TD b represents the reference drum torque, which is the reference value of the drum torque when the coal unloader operates normally, FD s represents the current conveyor belt tension, which is the tension value of the conveyor belt of the coal unloader in the current state, FD b represents the reference conveyor belt tension, which is the reference value of the conveyor belt tension when the coal unloader operates normally;

[0119] Call the drum torque data, the current drum torque TD mIt can be obtained through real-time monitoring. For example, the measured value of the drum torque of a coal unloader in an abnormal state is 850 N·m, while the reference drum torque TD b is 800 N·m. At the same time, the conveyor belt tension data is called, and the current conveyor belt tension FD s The measured value is 1100 N, and the reference conveyor belt tension FD b is 1000 N. Substitute the values into the formula for calculation:

[0120]

[0121] The calculated load change rate LD = 0.0336. If the interval of the medium load state is [0.03, 0.05], comparing with it, this result indicates that the current coal unloader is in the medium load state, which can be used as the basis for subsequent adjustment of the operation mode. Specifically, it is necessary to count the various operating conditions of the coal unloader for 6 months or 1 year, and then according to the data distribution, use the normal distribution analysis method to determine the typical value range of LD in different states. For example, take the 90% confidence interval in each state, and set the LD intervals of different abnormal states as the normal state (LD ≤ 0.02), mild abnormality (0.02 < LD ≤ 0.03), moderate abnormality (0.03 < LD ≤ 0.05), and severe abnormality (LD > 0.05).

[0122] The adjustment parameter screening sub-module compares based on the current working load level with the adjustment requirements of the abnormal state determination result, screens the response operation parameters of the coal unloader that meet the current state, and adjusts the operation mode of the coal unloader to obtain the response adjustment parameter set;

[0123] Call the set of coal unloader operation parameters that meet this load level. For example, in the medium load state, it includes adjustment parameters such as reducing the conveyor belt speed and reducing the change range of the drum torque. First, compare with the adjustment requirements of the abnormal state determination result. For example, the adjustment range of the drum torque is set within 10 N·m, and the adjustment range of the conveyor belt tension is set within 20%. Screen the response operation parameters of the coal unloader that meet the current state. For example, adjust the conveyor belt speed to 90% of the original operating speed, limit the change range of the drum torque within 8 N·m, and adjust the operation mode of the coal unloader to obtain the response adjustment parameter set.

[0124] Please refer to Figure 6 , the fault prediction module includes:

[0125] The change rate analysis sub-module, based on the response adjustment parameter set, obtains the data sequences of the hydraulic system pressure, the temperature of the bucket wheel drive motor, and the conveyor belt tension, calculates the time change rate of the data at adjacent time points, screens the abnormal fluctuation points, and obtains the parameter change rate after screening;

[0126] First, obtain the real-time data of the hydraulic system pressure, the temperature of the bucket wheel drive motor, and the conveyor belt tension. The data is collected by sensors, and each data point is timestamped. To ensure the continuity and accuracy of the data, time alignment processing is required for the data. For example, take the average value of multiple data points within one minute to obtain stable time-series data. Next, calculate the change rate of the data at adjacent time points, that is, calculate the numerical difference between the current time point and the previous time point, and divide it by the time interval to obtain the change rate value. In practical applications, for example, if the temperature of the bucket wheel drive motor rises from 60°C to 65°C within 5 minutes, then its temperature change rate is This calculation method is also applicable to the calculation of the change rates of the hydraulic system pressure and the conveyor belt tension. After calculating the change rates of all data points, it is necessary to screen out abnormal fluctuation points, set a reference change rate threshold, and determine the criteria for abnormal fluctuation points. For example, under normal operating conditions, the temperature change rate of the bucket wheel drive motor is generally less than 0.5°C / min. If the rate value calculated at a certain moment is greater than 2°C / min, it can be regarded as an abnormal fluctuation point and needs to be excluded or further analyzed. During the screening process, compare the change rate trends in the previous and subsequent time periods, and exclude instantaneous peaks or invalid fluctuation points to ensure that the obtained change rate data can truly reflect the equipment status and obtain the change rate of the screened parameters.

[0127] Based on the change rate of the screened parameters, the key factor screening sub-module analyzes the change trends of each parameter before and after the occurrence of abnormal conditions, calculates the fluctuation amplitude under different states, compares it with the rate difference, and screens out the parameters with key fluctuation ranges to form a set of fault impact factors;

[0128] First, take the data in the normal operating state as a reference benchmark, establish a historical data window. For example, take the operating data in the past 7 days, calculate the change range of each parameter, and determine its normal fluctuation range. In actual operation, for example, the temperature of the bucket wheel drive motor fluctuates within the range of 55°C to 65°C under normal conditions. If the temperature rises rapidly from 62°C to 72°C before a certain anomaly occurs and the change trend is significant, it indicates that this parameter has an obvious fluctuation before the fault. To quantify this change trend, it is necessary to calculate the fluctuation amplitude in the different states, that is, calculate the average change rate in the normal state and the abnormal state and find their difference. For example, the average temperature change rate in the normal state is 0.3°C / min, and the average temperature change rate in the abnormal state is 1.5°C / min, then the fluctuation amplitude is 1.5 - 0.3 = 1.2°C / min. Further, compare the rate differences to determine which parameters have the most significant impact on the abnormal state, and screen out the key parameters with a large fluctuation range to form a fault impact factor set. For example, if the change rate difference of the hydraulic system pressure is 0.2 MPa / min, the change rate difference of the conveyor belt tension is 0.15 kN / min, and the change rate difference of the bucket wheel drive motor temperature reaches 1.2°C / min, it shows that the fluctuation of the bucket wheel drive motor temperature has a greater impact on the fault. Therefore, it is included in the fault impact factor set.

[0129] Based on the fault impact factor set, the fault trend prediction sub-module calculates the fluctuation range of the impact factors in the historical data, using the formula:

[0130]

[0131] Predict the change amount of the fault trend in the future time period to obtain the fault trend prediction result. Among them, TB f represents the fault trend prediction value, indicating the change of the fault trend in the future time period. WB j is the weight of the j-th impact factor, indicating the relative importance of this impact factor in the fault prediction. VB j represents the change rate of the j-th impact factor, and the calculation method is the change rate of this parameter at adjacent time points. RB j represents the maximum value of the j-th impact factor in the historical data. MB j represents the minimum value of the j-th impact factor in the historical data. n B is the number of impact factors;

[0132] If the change rate of the bucket wheel drive motor temperature VB 1 = 1.2°C / min, the historical maximum value RB 1 = 75°C, the minimum value MB 1 = 55°C, the impact factor weight WB 1 = 0.5, the change rate of the conveyor belt tension VB 2= 0.15 kN / min, historical maximum value RB 2 = 10 kN, minimum value MB 2 = 7 kN, influence factor weight WB 2 = 0.3, hydraulic system pressure change rate VB 3 = 0.2 MPa / min, historical maximum value RB 3 = 12 MPa, minimum value MB 3 = 8 MPa, influence factor weight WB 3 = 0.2, substitute into the calculation formula:

[0133]

[0134] Calculate step by step to get:

[0135]

[0136] The predicted change in the fault trend within the future time period is 0.1713. Screen the moments when the change in the fault trend exceeds the warning threshold. If it exceeds the warning threshold, early warning or adjustment measures need to be taken to obtain the fault trend prediction result.

[0137] Regarding the warning threshold, it is necessary to compare the trend changes of the equipment during the fault-free and pre-fault periods, extract the distribution of historical fault trend prediction values within a certain period (such as the past 6 months), and calculate its mean μ and standard deviation σ;

[0138] Use the standard deviation range method (z-score method) to set the preliminary threshold. Generally speaking, the TB f value during normal operation of the equipment should be concentrated around μ, while the TB f usually deviates significantly from the mean before a fault. Therefore, it can be set that:

[0139] Ordinary warning threshold: TB warn = μ + 2σ (i.e., 2 times the standard deviation range, capturing 95.4% of the normal data);

[0140] Emergency warning threshold: TB critical = μ + 3σ (i.e., 3 times the standard deviation range, capturing 99.7% of the normal data, and if it exceeds this range, a fault is highly suspected);

[0141] It is also necessary to make adjustments in combination with the equipment operating conditions and industry standards. For example, the historical data analysis of a certain bucket wheel reclaimer shows that:

[0142] During normal operation of the equipment, TB f Average value μ = 0.08, standard deviation σ = 0.03;

[0143] 2 hours before the equipment fails, TB fThe mean value is approximately 0.15 - 0.18;

[0144] Thirty minutes before the equipment fails, TB f Suddenly rises above 0.25;

[0145] Based on this data:

[0146] Set the normal warning threshold: TB warn = 0.08 + 2 × 0.03 = 0.14;

[0147] Set the emergency warning threshold: TB critical = 0.08 + 3 × 0.03 = 0.17;

[0148] To avoid false alarms, it is necessary to monitor the warning trigger rate within a period of time. If it is found that the threshold is too low and causes too many false alarms, the warning threshold can be appropriately increased. For example, if there are more than 5 false alarms per week under normal operating conditions, adjust the normal warning threshold to 0.15 and observe the subsequent operation. In addition, the sliding window method can be used to recalculate the mean and standard deviation every week and dynamically adjust the threshold to ensure that the system adapts to the long-term change trend of the equipment.

[0149] An intelligent monitoring method for a coal unloader, comprising the following steps:

[0150] S1: Based on the cloud platform, monitor the operation data of the coal unloader, analyze the single-point information in the data stream, analyze the fluctuation range of the data within the time series, eliminate the timestamp fault data, and obtain the status monitoring data set;

[0151] S2: Based on the status monitoring data set, detect the changes in the drum torque and the temperature of the bucket wheel drive motor, calculate the change rate within the time series, mark the data with the change rate exceeding the limit, and adjust its transmission priority in the data stream to obtain the priority adjusted data stream;

[0152] S3: Based on the priority adjusted data stream, calculate the offset of each parameter within the normal operation range, screen the parameters with critical offset degree, and evaluate the abnormal state in combination with the abnormal occurrence duration to obtain the abnormal score index of the coal unloader;

[0153] S4: Based on the abnormal score index of the coal unloader, compare the changes when the abnormality occurs, determine the current working load level according to the load change, screen the response parameters that meet the current load level, and adjust the operation mode of the coal unloader to obtain the response adjusted parameter set;

[0154] S5: Based on the response adjusted parameter set, analyze the parameter change rate before and after the abnormal state occurs, screen the parameters with critical change rate, analyze the fluctuation range of the influencing factors in the historical data, and predict the change trend in the future time period to obtain the fault trend prediction result.

[0155] The above are only the preferred embodiments of the present invention, and do not limit the present invention in other forms. Any person skilled in the art may use the technical content disclosed above to make changes or modifications into equivalent embodiments with equivalent changes and apply them to other fields. However, as long as it does not depart from the technical solution content of the present invention, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

1. An intelligent monitoring system for a coal unloader, characterized in that: The system comprises: The operation data acquisition module is based on the cloud platform to monitor the operation data of the coal unloader, analyze the fluctuation range of the data in the time series, remove the timestamp fault data, and obtain the status monitoring data set; The priority scheduling module detects the change of the drum torque and the bucket wheel drive motor temperature based on the state monitoring data set, marks the data with rate change exceeding the limit, adjusts its transmission priority in the data stream, and obtains the priority adjustment data stream; The abnormal state detection module adjusts the data stream based on the priority, calculates the offset of each parameter, selects the key offset parameters, evaluates the abnormal state in combination with the duration of the abnormality, and obtains the coal unloader abnormality score index; The abnormal state response module compares the changes when the abnormality occurs based on the coal unloader abnormality scoring index, determines the current workload level according to the load change, selects the response parameters that meet the current workload level, adjusts the coal unloader operation mode, and obtains a response adjustment parameter set; The fault prediction module analyzes the parameter change rate before and after the abnormal state occurs based on the response adjustment parameter set, selects the key parameters of the change rate, predicts the change trend in the future time period, and obtains the fault trend prediction result.

2. The intelligent monitoring system for coal unloader according to claim 1, characterized in that: The state monitoring data set includes torque data, temperature data, tension data, and pressure data; the priority adjustment data stream includes change rate marking information, data fluctuation screening results, and priority adjustment results; the coal unloader abnormality scoring indicators include offset calculation results, error value reconstruction information, abnormality scores, and duration analysis results; the response adjustment parameter set includes state level analysis results, load level determination results, and operation mode adjustment results.

3. The intelligent monitoring system for coal unloader according to claim 1, characterized in that: The operation data acquisition module comprises: The data analysis submodule is based on the cloud platform and collects data on drum torque, bucket wheel drive motor temperature, conveyor belt tension, and hydraulic system pressure. It analyzes single-point information in the data stream, extracts measurement values ​​at multiple time points, and constructs a time series data set. The time series fluctuation analysis submodule analyzes the fluctuation range of each parameter on the time axis based on the time series data set, calculates the numerical difference between consecutive time points, detects abnormal mutation points, and uses the formula: Calculate the mutation point volatility index V t , filter the points where data fluctuations occur suddenly, and obtain the data fluctuation mutation points, where X t represents the measured value at time point t, X t-1 represents the measured value at time point t-1, represents the mean of the time series data, and N represents the total number of data points in the time series; The state monitoring data screening submodule removes the timestamp fault data based on the data fluctuation mutation points, screens the key data intervals of continuity, and organizes the screened data to obtain the state monitoring data set.

4. The intelligent monitoring system for coal unloader according to claim 1, characterized in that: The priority scheduling module comprises: The state monitoring submodule detects the drum torque and bucket wheel drive motor temperature data based on the state monitoring data set, calculates the incremental values ​​of multiple time points in the time series, and obtains the time series incremental data; The rate analysis submodule calculates the rate of change within the time series based on the incremental data of the time series, using the formula: Compared with the normal fluctuation range, the single data with rate changes exceeding the range are screened to obtain the over-limit rate data, among which RS t Represents the rate change value, TS i Represents the bucket wheel drive motor temperature at time i, MS i Represents the value of the drum torque at time i, n s Represents the number of sample points in the time series, TS i-1 represents the bucket wheel drive motor temperature at time i-1, MS i-1 represents the value of the drum torque at time i-1; The priority adjustment submodule calls the over-limit rate data, performs a change trend analysis, determines whether there is a continuous growth or abnormal fluctuation, and adjusts its transmission priority in the data stream to obtain a priority-adjusted data stream.

5. The intelligent monitoring system for coal unloader according to claim 1, characterized in that: The abnormal state detection module comprises: The parameter offset calculation submodule adjusts the data flow based on the priority, collects real-time data of conveyor belt tension, hydraulic system pressure and bucket wheel drive motor temperature, and calculates the offset amplitude relative to the historical baseline value using the formula: Get the parameter offset amplitude, where ΔZP r Represents the offset amplitude of the rth parameter, ZX r represents the measured value of the rth parameter, ZX b，r Represents the historical benchmark value of the rth parameter; The key parameter screening submodule screens the parameters with key offset degree based on the parameter offset amplitude, calculates the reconstruction error value thereof, determines whether it belongs to the abnormal range, and obtains the key offset parameters; The abnormality score calculation submodule performs abnormality score calculation based on the key offset parameters, and evaluates the abnormal state in combination with the duration of the abnormality to obtain the abnormality score index of the coal unloader.

6. The intelligent monitoring system for coal unloader according to claim 1, characterized in that: The abnormal state response module includes: The abnormal level determination submodule determines the abnormal state level based on the abnormal scoring index of the coal unloader, compares the torque fluctuation amplitude and tension force change trend under the abnormal level, and selects the corresponding torque change range and tension force adjustment amplitude to obtain the abnormal state determination result; The load change analysis submodule calculates the load change rate of the coal unloader during operation based on the abnormal state judgment result and the drum torque and conveyor belt tension data when the abnormality occurs, using the formula: Filter the load change interval that meets the abnormal level to obtain the current workload level, where LD represents the load change rate and TD represents the load change rate. m Represents the current reel torque, TD b Represents the reference drum torque, FD s Represents the current conveyor belt tension, FD b Represents the base conveyor belt tension; The adjustment parameter screening submodule compares the current workload level with the adjustment requirements of the abnormal state determination result, screens the coal unloader response operation parameters that meet the current state, and adjusts the coal unloader operation mode to obtain a response adjustment parameter set.

7. The intelligent monitoring system for coal unloader according to claim 1, characterized in that: The fault prediction module comprises: The change rate analysis submodule obtains the data sequence of hydraulic system pressure, bucket wheel drive motor temperature and conveyor belt tension based on the response adjustment parameter set, calculates the time change rate of the data at adjacent time points, screens abnormal fluctuation points, and obtains the parameter change rate after screening; The key factor screening submodule analyzes the change trend of each parameter before and after the abnormal state occurs based on the change rate of the screened parameters, calculates the fluctuation range under the difference state, compares it with the rate difference, screens the key parameters of the fluctuation range, and forms a set of fault influencing factors; The fault trend prediction submodule calculates the fluctuation range of the influencing factors in the historical data based on the fault influencing factor set, using the formula: Predict the change in fault trend in the future time period and obtain the fault trend prediction result, where TB f Represents the fault trend prediction value, WB j The weight of the jth impact factor, VB j Represents the rate of change of the jth influencing factor, RB j Represents the maximum value of the jth impact factor in historical data, MB j represents the minimum value of the jth impact factor in historical data, n B is the number of impact factors.

8. An intelligent monitoring method for a coal unloader, characterized in that: The intelligent monitoring system for a coal unloader according to any one of claims 1 to 7 comprises the following steps: S1: Based on the cloud platform, monitor the operation data of the coal unloader, parse the single point information in the data stream, analyze the fluctuation range of the data in the time series, remove the timestamp fault data, and obtain the status monitoring data set; S2: Based on the state monitoring data set, detect the change of the drum torque and the bucket wheel drive motor temperature, calculate the change rate in the time series, mark the data with rate change exceeding the limit, adjust its transmission priority in the data stream, and obtain the priority adjustment data stream; S3: Based on the priority adjustment data stream, calculate the offset of each parameter within the normal operating range, select the key parameters of the offset degree, evaluate the abnormal state in combination with the abnormality occurrence duration, and obtain the coal unloader abnormality score index; S4: Based on the coal unloader abnormality scoring index, compare the changes when the abnormality occurs, determine the current workload level according to the load change, select response parameters that meet the current workload level, adjust the coal unloader operation mode, and obtain a response adjustment parameter set; S5: Based on the response adjustment parameter set, analyze the parameter change rate before and after the abnormal state occurs, screen the key parameters of the change rate, analyze the fluctuation range of the influencing factors in the historical data, predict the change trend in the future time period, and obtain the fault trend prediction result.

Images