Fault monitoring and early warning system for scraper conveyor chain drive system

Abstract

This invention relates to a fault monitoring and early warning system for a scraper conveyor chain drive system, belonging to the field of scraper conveyor fault monitoring technology. It includes a data acquisition unit, a deep learning unit, a data processing unit, an alarm unit, and a display unit. The data acquisition unit includes an incremental encoder, a current transformer, an accelerometer, a tension sensor, and a wireless data acquisition device. The incremental encoder, current transformer, accelerometer, and tension sensor are all connected to the signal input terminal of the wireless data acquisition device. The signal output terminal of the wireless data acquisition device is connected to the signal input terminal of the deep learning unit. The first signal output terminal of the deep learning unit is connected to the signal input terminal of the data processing unit. The signal output terminal of the data processing unit is connected to the signal input terminals of both the alarm unit and the display unit. This invention not only provides more reliable monitoring and early warning results but also saves labor costs and has a high degree of intelligence.

Description

Technical Field

[0001] This invention relates to the field of scraper conveyor fault monitoring technology, and in particular to a fault monitoring and early warning system for a scraper conveyor chain drive system. Background Technology

[0002] Scraper conveyors, as effective traction devices, have long been widely used in mine production, serving as a crucial guarantee for the efficient and high-quality operation of underground fully mechanized mining and tunneling. Equipment failure can severely impact the continuous and efficient operation of the mine. Statistics show that over 50% of scraper conveyor failures are caused by malfunctions in their chain drive system. Common scraper conveyor chain drive system failures mainly include scraper chain jamming and scraper chain breakage. Considering adverse working conditions, the main factors affecting the reliability of the chain drive system include: uneven loads on the chute, excessively long conveyor belts, and the complex interactions within the transmission system. Since tension characteristics can effectively reflect the working state of the transmission system, studying the tension changes of the scraper chain can enable status monitoring and abnormal condition identification of the chain-driven transmission system.

[0003] Currently, there are many designs for fault detection devices for scraper conveyors. Patent CN102491067B uses a scraper to drive the idler roller, which in turn moves the sensor. However, this design uses a contact-type sensor with a single data type, resulting in low device reliability. Patent CN103434816B uses an inductive force sensor, speed sensor, current transformer, pressure sensor, and stroke sensor to measure five variables. While this provides a rich set of variables, the measurement results require manual interpretation, leading to high labor costs and low levels of automation. Previous research on scraper chain tension has largely focused on the direct measurement and analysis of tension parameters. However, because the tension sensor needs to be installed on the scraper chain and continuously move with it during operation, the working environment of the scraper conveyor significantly affects the measurement results, leading to inaccurate monitoring. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides a fault monitoring and early warning system for the chain drive system of a scraper conveyor. The technical solution of this invention is as follows:

[0005] A fault monitoring and early warning system for a scraper conveyor chain drive system includes a data acquisition unit, a deep learning unit, a data processing unit, an alarm unit, and a display unit. The data acquisition unit includes an incremental encoder, a current transformer, an acceleration sensor, a tension sensor, and a wireless data acquisition device. The incremental encoder, current transformer, acceleration sensor, and tension sensor are all connected to the signal input terminal of the wireless data acquisition device. The signal output terminal of the wireless data acquisition device is connected to the signal input terminal of the deep learning unit. The first signal output terminal of the deep learning unit is connected to the signal input terminal of the data processing unit. The signal output terminal of the data processing unit is connected to the signal input terminal of both the alarm unit and the first signal input terminal of the display unit.

[0006] The incremental encoder is used to detect the rotational speed signal of the drive motor of the scraper conveyor when it is working, the current transformer is used to detect the working current signal of the drive motor, the acceleration sensor is used to detect the horizontal vibration signal of the scraper conveyor, and the tension sensor is used to detect the scraper chain tension signal along the running direction of the scraper conveyor in the initial stage of the scraper conveyor's start-up operation.

[0007] The wireless data acquisition device is used to receive rotation speed signals, operating current signals, horizontal vibration signals, and scraper chain tension signals and then transmit them to the deep learning unit.

[0008] The deep learning unit is used to predict the tension of the scraper chain based on the rotational speed signal, the working current signal and the horizontal vibration signal, and obtain three sets of tension predictions. The three sets of tension predictions and the scraper chain tension signal are then transmitted to the data processing unit.

[0009] The data processing unit is used to determine the tension prediction value based on three sets of tension prediction values, and to determine whether the scraper chain of the scraper conveyor has malfunctioned based on the tension prediction value and the scraper chain tension signal. When it is determined that the scraper chain has malfunctioned, the unit sends the fault type to the alarm unit for alarm and sends the fault type to the display unit for display.

[0010] Optionally, the second signal output terminal of the deep learning unit is also connected to the second signal input terminal of the display unit;

[0011] The deep learning unit is used to plot the corresponding change curves based on the rotation speed signal, working current signal, horizontal vibration signal and scraper chain tension signal sent by the wireless data acquisition device in each sampling period.

[0012] The display unit is also used to display the change curves corresponding to the rotation speed signal, working current signal, horizontal vibration signal, and scraper chain tension signal.

[0013] Optionally, the incremental encoder is installed at the end of the rotating shaft of the drive motor of the scraper conveyor; the current transformer is installed at the inlet of the tail motor of the scraper conveyor; the acceleration sensor is installed on the upper surface of the scraper of the scraper conveyor; the tension sensor is fixed on the vertical chain of the scraper conveyor; and the wireless data acquisition device is installed on the chute of the scraper conveyor via a base.

[0014] Optionally, when the deep learning unit predicts the scraper chain tension based on the rotational speed signal, working current signal, and horizontal vibration signal, it first normalizes the rotational speed signal, working current signal, horizontal vibration signal, and scraper chain tension signal. Then, it inputs the normalized rotational speed signal, working current signal, and horizontal vibration signal into a pre-trained convolutional neural network. Based on the output of the convolutional neural network, it determines the tension prediction amount corresponding to the rotational speed signal, working current signal, and horizontal vibration signal, respectively.

[0015] Optionally, the data processing unit determines the tension prediction value based on the three sets of tension prediction values ​​through the following steps:

[0016] First, calculate the absolute error R between the predicted tension for each group and the tension signal of the scraper chain. i =abs(T i -T0), (i = 1, 2, 3), T i T0 represents the tension prediction corresponding to the rotational speed signal, working current signal, and horizontal vibration signal; T0 represents the scraper chain tension signal.

[0017] Then, define the weighting coefficients for each group of tension predictions as 0 < δ. i <1, (i = 1, 2, 3), based on population variance D(R i The variance of the absolute error of each group is represented by δ, which is calculated by the following set of equations to minimize the tension prediction of each group. i

[0018]

[0019] Finally, the predicted tension value T is determined using the formula T = δ1T1 + δ2T2 + δ3T3.

[0020] Optionally, when the data processing unit determines whether the scraper chain of the scraper conveyor has malfunctioned based on the tension prediction value and the scraper chain tension signal, it compares the tension prediction value and the scraper chain tension signal, and determines whether the scraper chain has been subjected to a large load impact, has experienced a chain jamming fault, or has experienced a chain breakage fault based on the comparison result.

[0021] All of the above-mentioned optional technical solutions can be combined arbitrarily, and the present invention will not provide a detailed description of the structure after each combination.

[0022] By means of the above solution, the beneficial effects of the present invention are as follows:

[0023] By setting up a data acquisition unit, a deep learning unit, a data processing unit, an alarm unit, and a display unit, the data acquisition unit includes an incremental encoder, a current transformer, an acceleration sensor, a tension sensor, and a wireless data acquisition device. Using the rotational speed signal, working current signal, and horizontal vibration signal of the scraper when the drive motor is working as references, the operating status of the scraper conveyor scraper chain is jointly determined. This not only makes the monitoring and early warning results more reliable, but also saves labor costs and has a high degree of intelligence.

[0024] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, the preferred embodiments of the present invention are described in detail below with reference to the accompanying drawings. Attached Figure Description

[0025] Figure 1 This is a schematic diagram of the system composition structure of the present invention.

[0026] Figure 2 This is a schematic diagram of the change curve of the predicted tension value under heavy load impact in an embodiment of the present invention.

[0027] Figure 3 This is a schematic diagram of the change curve of the predicted tension value when the chain fails in an embodiment of the present invention.

[0028] Figure 4 This is a schematic diagram of the change curve of the predicted tension value during a chain breakage failure in an embodiment of the present invention. Detailed Implementation

[0029] The specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and are not intended to limit the scope of the invention.

[0030] like Figure 1As shown, the fault monitoring and early warning system for the chain drive system of the scraper conveyor provided by the present invention includes a data acquisition unit, a deep learning unit, a data processing unit, an alarm unit, and a display unit. The data acquisition unit includes an incremental encoder, a current transformer, an acceleration sensor, a tension sensor, and a wireless data acquisition device. The incremental encoder, current transformer, acceleration sensor, and tension sensor are all connected to the signal input terminal of the wireless data acquisition device. The signal output terminal of the wireless data acquisition device is connected to the signal input terminal of the deep learning unit. The first signal output terminal of the deep learning unit is connected to the signal input terminal of the data processing unit. The signal output terminal of the data processing unit is connected to the signal input terminal of the alarm unit and the first signal input terminal of the display unit.

[0031] The incremental encoder is used to detect the rotational speed signal of the drive motor of the scraper conveyor when it is working, the current transformer is used to detect the working current signal of the drive motor, the acceleration sensor is used to detect the horizontal vibration signal of the scraper conveyor, and the tension sensor is used to detect the scraper chain tension signal along the running direction of the scraper conveyor in the initial stage of the scraper conveyor's start-up operation.

[0032] The wireless data acquisition device is used to receive rotation speed signals, operating current signals, horizontal vibration signals, and scraper chain tension signals and then transmit them to the deep learning unit.

[0033] The deep learning unit is used to predict the tension of the scraper chain based on the rotational speed signal, the working current signal and the horizontal vibration signal, and obtain three sets of tension predictions. The three sets of tension predictions and the scraper chain tension signal are then transmitted to the data processing unit.

[0034] The data processing unit is used to determine the tension prediction value based on three sets of tension prediction values, and to determine whether the scraper chain of the scraper conveyor has malfunctioned based on the tension prediction value and the scraper chain tension signal. When it is determined that the scraper chain has malfunctioned, the unit sends the fault type to the alarm unit to trigger an alarm to remind on-site personnel to perform equipment maintenance, and sends the fault type to the display unit for display.

[0035] Specifically, embodiments of the present invention may employ a dual-CPU design, with the deep learning unit and the data processing unit respectively configured with microprocessor one and microprocessor two, thereby improving the speed of data processing and ensuring the real-time performance of the fault monitoring and early warning system.

[0036] The deep learning unit and the data processing unit can communicate via SPI circuit; the deep learning unit communicates with each sensor of the data acquisition unit via RS-485; and the data processing unit communicates with the display unit and the alarm unit via RS-485.

[0037] Optionally, the second signal output terminal of the deep learning unit is also connected to the second signal input terminal of the display unit; the deep learning unit is used to plot corresponding change curves based on the rotational speed signal, operating current signal, horizontal vibration signal, and scraper chain tension signal sent by the wireless data acquisition device in each sampling period; the display unit is also used to display the change curves corresponding to the rotational speed signal, operating current signal, horizontal vibration signal, and scraper chain tension signal. The sampling frequencies of the incremental encoder, current transformer, accelerometer, and tension sensor are all set to 1kHz.

[0038] Specifically, the incremental encoder is installed at the end of the rotating shaft of the scraper conveyor's drive motor. The current transformer is installed at the tail motor inlet of the scraper conveyor, obtaining the operating current of the drive motor by detecting the current in the power line. The acceleration sensor is installed on the upper surface of the scraper of the scraper conveyor. The tension sensor is fixed on the vertical chain of the scraper conveyor. The wireless data acquisition device is mounted on the chute of the scraper conveyor via a base.

[0039] Optionally, when the deep learning unit predicts the scraper chain tension based on the rotational speed signal, working current signal, and horizontal vibration signal, it first normalizes the rotational speed signal, working current signal, horizontal vibration signal, and scraper chain tension signal. Then, it inputs the normalized rotational speed signal, working current signal, and horizontal vibration signal into a pre-trained convolutional neural network. Based on the output of the convolutional neural network, it determines the tension prediction amount corresponding to the rotational speed signal, working current signal, and horizontal vibration signal, respectively.

[0040] Since tension characteristics can effectively reflect the working state of a chain drive system, this invention aims to monitor the state of the chain drive system and identify abnormal operating conditions by studying the tension changes of the scraper chain. Specifically, the convolutional neural network model selected in this invention is a one-dimensional convolutional neural network (1DCNN). Because different signals can reflect the tension information of the scraper chain from different perspectives, to achieve uniformity among these signals, the deep learning unit in this invention first normalizes the four different signals collected by the four sensors. Then, the normalized rotational speed signal, working current signal, and horizontal vibration signal are used as the input vector of the convolutional neural network, and the tension prediction is used as the output vector. For the input vector, the first 600 data points are used as the training set, and the last 400 data points are used as the test set to verify the prediction results of the convolutional neural network.

[0041] The 1DCNN used has three layers, including a fully connected layer and a Softmax layer, three convolutional layers, and a pooling layer. Max pooling is selected for the pooling layer. The parameter settings for the convolutional and pooling layers are shown in Table 1. The normalized rotational speed signal, working current signal, and horizontal vibration signal are processed by the convolutional neural network to obtain three sets of tension predictions, denoted as T1, T2, and T3, which represent the tension predictions of the scraper chain based on the rotational speed signal, working current signal, and horizontal vibration signal, respectively. In addition, the scraper chain tension signal is denoted as T0, and T0, T1, T2, and T3 are input together into the data processing unit.

[0042] Table 1

[0043]

[0044] Furthermore, based on the theory of multi-sensor information fusion, a suitable BPA can be constructed to integrate information from different data sources, thereby eliminating data redundancy and obtaining accurate results. In this embodiment of the invention, the data processing unit uses data processed by a convolutional neural network to construct the BPA. Since different types of data have different relationships with the tension of the scraper chain, weighting coefficients are introduced to represent the degree of influence of different types of data on the tension. Specifically, the data processing unit determines the tension prediction value based on three sets of tension prediction values ​​through the following steps:

[0045] First, calculate the absolute error R between the predicted tension for each group and the tension signal of the scraper chain. i =abs(T i -T0), (i = 1, 2, 3), T i T0 represents the tension prediction corresponding to the rotational speed signal, working current signal, and horizontal vibration signal; T0 represents the scraper chain tension signal.

[0046] Then, define the weighting coefficients for each group of tension predictions as 0 < δ. i <1, (i = 1, 2, 3), based on the overall variance of tension D(R i The variance of the absolute error of each group is represented by δ, which is calculated by the following set of equations to minimize the tension prediction of each group. i

[0047]

[0048] Finally, the predicted tension value T is determined using the formula T = δ1T1 + δ2T2 + δ3T3.

[0049] Furthermore, the deep learning unit can combine tension prediction values ​​from different sampling periods to obtain a tension prediction curve and send it to the display unit, which then displays the tension prediction curve.

[0050] Optionally, when the data processing unit determines whether the scraper chain of the scraper conveyor has malfunctioned based on the tension prediction value and the scraper chain tension signal, it compares the tension prediction value T with the scraper chain tension signal T0, and determines whether the scraper chain has been subjected to a large load impact, experienced a chain jamming fault, or experienced a chain breakage fault based on the comparison result.

[0051] Specifically, when the scraper chain is subjected to a large load impact, T increases rapidly and is accompanied by high-frequency oscillation characteristics, with its peak value reaching up to 5.4 times T0. Figure 2 As shown. When a chain failure occurs, T abruptly changes and stabilizes at a very high level, with a peak value reaching 7.2 times T0, as... Figure 3 As shown. When a chain breakage occurs, the tension disappears instantly, and T becomes 0, as... Figure 4 As shown. In addition to identifying abnormal working conditions, embodiments of the present invention can also preset the tension threshold of the scraper chain. When the predicted tension value exceeds the tension threshold, it can be determined that the current coal conveying volume is too high. At this time, the diagnostic results can also be sent to the display unit for display and the alarm unit for alarm, so as to remind the on-site workers to appropriately reduce the coal mining speed of the coal mining machine.

[0052] In summary, the fault monitoring and early warning system for the chain drive system of the scraper conveyor proposed in this embodiment of the invention has the following characteristics:

[0053] 1. This invention uses the rotational speed signal of the drive motor, the operating current signal, and the horizontal vibration signal of the scraper as references to jointly determine the operating status of the scraper conveyor chain, thus overcoming the problem of low reliability caused by the single indicator in existing scraper conveyor monitoring systems. Furthermore, the scraper chain tension signal, which is most difficult to obtain in actual working conditions, is only measured at the initial stage of the scraper conveyor's operation and can be removed after a period of operation, reducing the impact of the working environment on the measurement results and making the monitoring and early warning results more accurate and reliable.

[0054] 2. This invention employs deep learning technology, leveraging the advantages of convolutional neural networks in parameter prediction, particularly in nonlinear and time-varying data, to determine the tension prediction amount. The prediction results are accurate, reliable, and highly intelligent.

[0055] 3. Based on the theory of multi-sensor data fusion, this invention introduces a weighting coefficient δ, taking into account the varying degrees of influence of different types of data on the tension prediction. i This makes the obtained tension prediction values ​​more accurate, and the monitoring and early warning results more accurate and reliable.

[0056] 4. The sensors selected in this invention are easy to install, and the data required for monitoring and early warning are easy to obtain.

[0057] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A fault monitoring and early warning system for a scraper conveyor chain drive system, characterized in that, It includes a data acquisition unit, a deep learning unit, a data processing unit, an alarm unit, and a display unit; the data acquisition unit includes an incremental encoder, a current transformer, an accelerometer, a tension sensor, and a wireless data acquisition device. The incremental encoder, current transformer, accelerometer, and tension sensor are all connected to the signal input terminal of the wireless data acquisition device. The signal output terminal of the wireless data acquisition device is connected to the signal input terminal of the deep learning unit. The first signal output terminal of the deep learning unit is connected to the signal input terminal of the data processing unit. The signal output terminal of the data processing unit is connected to the signal input terminal of the alarm unit and the first signal input terminal of the display unit. The incremental encoder is used to detect the rotational speed signal of the drive motor of the scraper conveyor when it is working, the current transformer is used to detect the working current signal of the drive motor, the acceleration sensor is used to detect the horizontal vibration signal of the scraper conveyor, and the tension sensor is used to detect the scraper chain tension signal along the running direction of the scraper conveyor in the initial stage of the scraper conveyor's start-up operation. The wireless data acquisition device is used to receive rotation speed signals, operating current signals, horizontal vibration signals, and scraper chain tension signals and then transmit them to the deep learning unit. The deep learning unit is used to predict the tension of the scraper chain based on the rotational speed signal, the working current signal and the horizontal vibration signal, and obtain three sets of tension predictions. The three sets of tension predictions and the scraper chain tension signal are then transmitted to the data processing unit. The data processing unit is used to determine the tension prediction value based on three sets of tension prediction values, and to determine whether the scraper chain of the scraper conveyor has malfunctioned based on the tension prediction value and the scraper chain tension signal. When it is determined that the scraper chain has malfunctioned, the unit sends the fault type to the alarm unit for alarm and sends the fault type to the display unit for display. The data processing unit determines the predicted tension value based on the three sets of predicted tension values ​​through the following steps: First, calculate the absolute error between the predicted tension for each group and the tension signal of the scraper chain. , T i This represents the tension prediction value corresponding to the rotational speed signal, operating current signal, and horizontal vibration signal, where, T 1 , T 2 , T 3 These represent the predicted tension of the scraper chain based on the rotational speed signal, the operating current signal, and the horizontal vibration signal, respectively. T 0 This indicates the tension signal of the scraper chain; Then, the weighting coefficients of each group of tension predictions are defined. According to the overall variance , The variance of the absolute error of each group is calculated by the following system of equations. The weighting coefficients of each group of tension predictions at the minimum ; ； Finally, through the formula Determine the predicted tension value T .

2. The fault monitoring and early warning system for the chain drive system of the scraper conveyor according to claim 1, characterized in that, The second signal output terminal of the deep learning unit is also connected to the second signal input terminal of the display unit; The deep learning unit is used to plot the corresponding change curves based on the rotation speed signal, working current signal, horizontal vibration signal and scraper chain tension signal sent by the wireless data acquisition device in each sampling period. The display unit is also used to display the change curves corresponding to the rotation speed signal, working current signal, horizontal vibration signal, and scraper chain tension signal.

3. The fault monitoring and early warning system for the chain drive system of the scraper conveyor according to claim 1, characterized in that, The incremental encoder is installed at the end of the rotating shaft of the drive motor of the scraper conveyor; the current transformer is installed at the inlet of the tail motor of the scraper conveyor; the acceleration sensor is installed on the upper surface of the scraper of the scraper conveyor; the tension sensor is fixed on the vertical chain of the scraper conveyor; and the wireless data acquisition device is installed on the chute of the scraper conveyor via a base.

4. The fault monitoring and early warning system for the chain drive system of the scraper conveyor according to claim 1, characterized in that, When the deep learning unit predicts the tension of the scraper chain based on the rotational speed signal, working current signal, and horizontal vibration signal, it first normalizes the rotational speed signal, working current signal, horizontal vibration signal, and scraper chain tension signal. Then, it inputs the normalized rotational speed signal, working current signal, and horizontal vibration signal into a pre-trained convolutional neural network. Based on the output of the convolutional neural network, it determines the tension prediction amount corresponding to the rotational speed signal, working current signal, and horizontal vibration signal, respectively.

5. The fault monitoring and early warning system for the chain drive system of the scraper conveyor according to claim 1, characterized in that, When the data processing unit determines whether the scraper chain of the scraper conveyor has malfunctioned based on the tension prediction value and the scraper chain tension signal, it compares the tension prediction value and the scraper chain tension signal, and determines whether the scraper chain has been subjected to a large load impact, has experienced a chain jamming fault, or has experienced a chain breakage fault based on the comparison result.

Images