Equipment health assessment method and device and electronic equipment

Abstract

The invention discloses an equipment health assessment method and device and electronic equipment. The method comprises the following steps: acquiring running state data of the equipment, the running state data being various parameter characteristics acquired at multiple positions of the equipment; adjusting the initial weights corresponding to the various parameter features to obtain target operation state data; a health assessment model is adopted to assess the target operation state data, an equipment health degree score is obtained, and the equipment health degree score is used for quantitatively assessing the equipment health state. The technical problem that the actual health state of the equipment cannot be accurately reflected due to the fact that the static weight factor is adopted to evaluate the health state of the equipment in the prior art is solved.

Description

Technical Field

[0001] The present application relates to the field of industrial control, and in particular to a method, apparatus and electronic device for equipment health assessment. Background Art

[0002] Industrial equipment usually relies on static weight factors to evaluate the health status of equipment. This fixed weight model is difficult to adapt to changes in the status and environment during equipment operation, and thus cannot accurately reflect the actual health status of the equipment under complex working conditions.

[0003] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0004] The embodiments of the present application provide a method, apparatus and electronic device for evaluating the health status of a device, so as to at least solve the technical problem that the related art uses static weight factors to evaluate the health status of the device and accurately reflect the actual health status of the device.

[0005] According to one aspect of an embodiment of the present application, a method for device health assessment is provided, comprising: obtaining operating status data of a device, wherein the operating status data comprises: multiple parameter characteristics collected at multiple locations of the device; adjusting initial weights corresponding to the multiple parameter characteristics to obtain target operating status data; and evaluating the target operating status data using a health assessment model to obtain a device health score, wherein the device health score is used to quantitatively evaluate the health status of the device.

[0006] In some embodiments of the present application, the initial weights corresponding to multiple parameter features are determined in the following manner: obtaining historical sample data of the device, wherein the historical sample data includes: historical operating status data of the device and a real health score corresponding to the historical operating status data; using a linear regression model to predict the historical operating status data to obtain a predicted health score; using a loss function to calculate the loss value between the predicted health score and the real health score, adjusting the parameters of the linear regression model according to the loss value, and obtaining the initial weights corresponding to the multiple parameters.

[0007] In some embodiments of the present application, initial weights corresponding to multiple parameter features are adjusted to obtain target operating status data, including: calculating a first health score corresponding to multiple parameter features with initial weights; using a preset model to predict multiple parameter features with initial weights to obtain a second health score corresponding to the multiple parameter features; calculating the error between the first health score and the second health score; adjusting the initial weights corresponding to the multiple parameter features according to the error to obtain target weights corresponding to the multiple parameter features; and determining the multiple parameter features with the target weights as the target operating status data.

[0008] In some embodiments of the present application, the initial weights corresponding to the various parameter features are adjusted according to the error to obtain the target weights corresponding to the various parameter features, including: the target weight is calculated by the following formula: β i (t+1)=β i (t)+α(e(t)-e(t-1))·χ i , where β i (t) is the weight of the i-th sensor at time t, α is the frequency of weight adjustment, e(t) is the evaluation error at time t, X i is the data collected by the i-th sensor.

[0009] In some embodiments of the present application, the multiple parameter features are determined in the following manner, including: collecting data of multiple parameters at a preset collection frequency to obtain multiple parameter data sets corresponding to the multiple parameters collected respectively at different preset time intervals; determining parameter types of the multiple parameters; for each parameter data set, determining the parameter type to which the parameter data in each parameter data set belongs; determining a processing strategy corresponding to the parameter type, wherein the processing strategy includes one of the following: extracting the maximum value, extracting the minimum value, calculating the average value, calculating the range, and calculating the variance; processing the multiple parameter data in each parameter data set according to the processing strategy to obtain parameter features corresponding to each parameter data set in the multiple parameter features.

[0010] In some embodiments of the present application, the method also includes: comparing the health score with a preset maintenance threshold and a preset warning threshold, respectively, wherein the preset maintenance threshold is higher than the preset warning threshold; generating maintenance suggestion information when the health score is lower than the preset maintenance threshold; and generating warning information when the health score is lower than the preset warning threshold.

[0011] In some embodiments of the present application, the initial weights corresponding to the various parameter features are adjusted respectively, and before the target operating status data is obtained, the method also includes: in response to a selection instruction, determining a target filtering method from a plurality of types of filtering methods, wherein the plurality of types of filtering methods include median filtering and mean filtering; using the target filtering method to remove noise of the various parameter features to obtain filtered data corresponding to the various parameter features; using a minimum-maximum value normalization algorithm to normalize the filtered data to obtain normalized data corresponding to the various parameter features.

[0012] According to another aspect of the embodiment of the present application, a device for evaluating equipment health is also provided, including: an acquisition module for acquiring operating status data of the equipment, wherein the operating status data is a plurality of parameter characteristics collected at multiple locations of the equipment; an adjustment module for adjusting initial weights corresponding to the plurality of parameter characteristics, respectively, to obtain target operating status data; and an evaluation module for evaluating the target operating status data using a health evaluation model to obtain an equipment health score, wherein the equipment health score is used to quantitatively evaluate the health status of the equipment.

[0013] According to another aspect of the embodiments of the present application, there is also provided an electronic device, including: a memory and a processor, the memory being used to store program instructions; the processor being connected to the memory and being used to execute the method for implementing the above-mentioned device health assessment.

[0014] According to another aspect of the embodiments of the present application, a computer program product is provided, including computer instructions, which implement the above-mentioned device health assessment method when executed by a processor.

[0015] In an embodiment of the present application, by adjusting the initial weights of multiple parameter features in the collected operating status data, and then using the health assessment model to perform a health assessment on the adjusted operating status data, the purpose of dynamically adjusting the weights of each parameter feature according to the actual working conditions during the operation of the equipment is achieved, thereby achieving the technical effect of accurately reflecting the actual health status of the equipment, and further solving the technical problem that the related technology uses static weight factors to evaluate the health status of the equipment and cannot accurately reflect the actual health status of the equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0017] Figure 1 It is a hardware structure block diagram of a computer terminal of a device health assessment method according to an embodiment of the present application;

[0018] Figure 2 is a flow chart of a device health assessment method according to an embodiment of the present application;

[0019] Figure 3 is a schematic diagram of a device health assessment method according to an embodiment of the present application;

[0020] Figure 4 It is a structural schematic diagram of a device for equipment health assessment according to an embodiment of the present application. DETAILED DESCRIPTION

[0021] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0022] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0023] In order to better understand the embodiments of the present application, the technical terms involved in the embodiments of the present application are explained as follows:

[0024] Parameter characteristics: refers to the specific numerical data collected by various sensors during the operation of the equipment. These data can reflect the status and performance of the equipment. Through a certain collection frequency and time interval, a continuous data stream of the equipment's operating status is collected. By analyzing these data, characteristic values ​​(i.e., parameter characteristics) can be extracted for the assessment of the equipment's health status.

[0025] Dynamic weight: refers to the process of assigning different importance weights to various quantitative parameter characteristics according to actual operating conditions and historical data during the equipment health assessment process. This weight is not fixed, but is dynamically adjusted according to the equipment's operating status, environmental changes, and historical performance.

[0026] Equipment health monitoring and evaluation is one of the key technologies in modern industry and manufacturing. With the development of industrial automation and intelligence, enterprises have higher and higher requirements for equipment reliability and maintenance. In related technologies, the multi-parameter evaluation method based on a single weight factor uses multiple sensor data for health evaluation, but the weights of each parameter are usually static and cannot be adjusted dynamically. This method may have a certain accuracy in the early stage, but as the equipment working conditions change and time passes, the static weights are difficult to adapt to new operating conditions, resulting in the evaluation results gradually becoming inaccurate; the evaluation model based on empirical rules uses expert knowledge and empirical rules for health evaluation, such as analyzing sensor data and judging equipment status based on rule engines. Although empirical rules can improve the accuracy of evaluation to a certain extent, the limitation of this method is that the formulation and updating of rules rely on expert knowledge and it is difficult to handle complex and changeable operating environments; some methods introduce machine learning models, such as linear regression or decision trees, for equipment health evaluation. The model learns the relationship between parameters through training data and is used to evaluate the health status. However, simple machine learning models often find it difficult to capture the complex dynamic relationships in equipment operation, and the generalization ability and robustness of the model are poor.

[0027] It can be seen that the current health status assessment of industrial equipment usually relies on the analysis of a single parameter or static weight factor. This traditional method often ignores the dynamic changes and mutual influence of various parameters during the operation of the equipment, resulting in the inability to accurately reflect the actual health status of the equipment under complex working conditions, and the accuracy and reliability of the assessment results are insufficient. In addition, the assessment of the health status of the equipment depends on regular manual inspection and maintenance, which is highly subjective and time-delayed. Although regular maintenance can avoid certain failures, it is difficult to predict the real-time health status of the equipment, resulting in high maintenance costs and a high risk of unplanned downtime. In general, the shortcomings of the relevant technologies are mainly reflected in the following aspects: in terms of data collection, the existing technology usually relies on a single or a few parameters for equipment health assessment, which cannot fully reflect the operating status of the equipment; in terms of weight allocation, the use of static weight factors lacks a dynamic adjustment mechanism, which makes it difficult to adapt to the changes in the importance of various parameters during the operation of the equipment; in terms of the accuracy of the assessment model, the traditional assessment method fails to make full use of multi-dimensional data and dynamic weight adjustment, resulting in low accuracy and reliability of the assessment results; in terms of real-time and early warning capabilities, the assessment methods in the relevant technologies usually have delays and cannot provide real-time assessment and early warning of the health status of the equipment in a timely manner. In order to solve the above technical problems, the embodiments of the present application provide corresponding solutions, which are described in detail below.

[0028] The method embodiment of device health assessment provided in the embodiment of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Figure 1The hardware structure block diagram of a computer terminal for implementing a method for device health assessment is shown. Figure 1 As shown, the computer terminal 10 may include one or more (102a, 102b, ..., 102n are used to illustrate) processors (the processor may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA), a memory 104 for storing data, and a transmission module 106 for communication functions connected via a wired and / or wireless network. In addition, it may also include: a display, a keyboard, a cursor control device, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of the I / O interface), a network interface, and a BUS bus. Those skilled in the art can understand that Figure 1 The structure shown is only for illustration and does not limit the structure of the above electronic device. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations shown.

[0029] It should be noted that the one or more processors and / or other data processing circuits described above may generally be referred to herein as "data processing circuits". The data processing circuits may be embodied in whole or in part as software, hardware, firmware, or any other combination thereof. In addition, the data processing circuit may be a single independent processing module, or may be incorporated in whole or in part into any of the other components in the computer terminal 10. As described in the embodiments of the present application, the data processing circuit acts as a processor control (e.g., selection of a variable resistor terminal path connected to an interface).

[0030] The memory 104 can be used to store software programs and modules of application software, such as program instructions / data storage devices corresponding to the method of device health assessment in the embodiment of the present application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory 104, that is, realizing the above-mentioned method of device health assessment. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely arranged relative to the processor, and these remote memories may be connected to the computer terminal 10 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0031] The transmission module 106 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of the computer terminal 10. In one example, the transmission module 106 includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission module 106 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0032] The display may be, for example, a touch screen liquid crystal display (LCD) that enables a user to interact with a user interface of the computer terminal 10 .

[0033] It should be noted that, in some optional embodiments, the above Figure 1 The computer terminal shown may include hardware elements (including circuits), software elements (including computer code stored on a computer-readable medium), or a combination of hardware elements and software elements. It should be noted that Figure 1 This is merely one example of a particular embodiment and is intended to illustrate the types of components that may be present in the computer terminal described above.

[0034] In the above operating environment, an embodiment of the present application provides a method embodiment of device health assessment. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0035] Figure 2 is a flow chart of a method for device health assessment according to an embodiment of the present application, such as Figure 2 As shown, the method comprises the following steps:

[0036] Step S202, obtaining the operating status data of the device, wherein the operating status data includes: multiple parameter characteristics collected at multiple locations of the device.

[0037] In the above step S202, the device is any device that can be operated or perform tasks, such as a generator in a power plant, a drilling rig on an oil drilling platform, etc. The operating status data is information collected from the device through sensors, instruments or other technical means. This information can be digital, text, image or other forms, used to describe the status or performance of the device. In some embodiments, the collected information can be pre-processed to obtain the operating status data. The multiple locations of the device are different parts or areas of the device. For example, a large industrial machine may install sensors in multiple key parts such as motors, bearings, control systems, etc. to monitor the status of different locations.

[0038] Parameter features are quantitative indicators used to describe the state of the equipment, such as physical quantities such as temperature, pressure, speed, current, rotation speed, sound voltage, etc., or more abstract indicators, such as the working cycle of the equipment, failure rate, etc. Multiple parameter features are multiple types of data collected from different locations of the equipment to fully understand the state of the equipment. For example, multiple sensors are used to simultaneously collect specific values ​​such as vibration, temperature, current, pressure, rotation speed, sound, etc. of the equipment. The collected data of multiple types can be used as parameter features, or a preset number of types of data can be selected as parameter features according to preset rules. For example, by calculating the correlation coefficient between different parameters, parameters that are highly correlated with the health status of the equipment are selected or the machine learning model (such as random forest, gradient boosting tree, etc.) is used to evaluate the contribution of each feature to the model's prediction ability, and features with high importance are selected.

[0039] In some embodiments, sensors are installed at multiple locations of the device to collect multiple parameter characteristics including vibration, temperature, current, pressure, rotation speed and sound. Multiple sensors collect data at regular intervals to form a continuous data stream of the device operation status. The collected raw data is recorded as X={χ1, χ2, ..., χ n}, where X i is the raw data collected by the i-th sensor.

[0040] Step S204, adjusting the initial weights corresponding to the various parameter features to obtain target operating state data.

[0041] In the above step S204, the initial weight is the weight of the above-mentioned multiple parameter characteristics at the time of collecting the operating status data, which is used to reflect the influence of the corresponding parameter characteristics on the health status of the equipment. The initial weight can be determined by at least one of historical data analysis, expert experience, machine learning algorithm and statistical benchmark method. In some embodiments, by analyzing the historical operating status data of the equipment, the influence of each parameter on the health status of the equipment under different operating conditions and environmental conditions is identified, and statistical analysis methods, such as correlation analysis, regression analysis, etc., are used to quantify the relationship between each parameter and the health status of the equipment, so as to determine its initial weight.

[0042] In order to avoid the problem of insufficient assessment accuracy caused by using fixed weights to perform health assessments on the parameters of each measuring point of the equipment, during the operation of the equipment, the initial weights can be dynamically adjusted using a preset model based on the real-time working conditions and historical data of the equipment. The preset model includes a machine learning model, a statistical analysis model, a reinforcement learning model, etc., and can also include a hybrid model composed of multiple models. In some embodiments, the initial weights are adjusted by real-time monitoring of the changes in the parameters of each measuring point, combining real-time data with an adaptive weighting algorithm. For example, by real-time monitoring of the temperature and vibration data of the motor, the initial weights of each parameter feature are adjusted so that it can reflect the state changes of the motor in a timely manner. In an optional embodiment, the target weights (i.e., the adjusted weights) of data such as temperature and vibration are calculated by the following formula:

[0043] β i (t+1)=β i (t)+α(e(t)-e(t-1))·χ i

[0044] Among them, β i (t) is the weight of the i-th sensor at time t, α is the frequency of weight adjustment, e(t) is the evaluation error at time t, X i is the data collected by the i-th sensor.

[0045] Through this weight adjustment formula, the weight of parameter features can be quickly adjusted according to the real-time changes in the equipment operating status, thereby improving the adaptability and accuracy of the model.

[0046] The above-mentioned target operating status data is operating status data with adjusted weights. Since the data can be used to evaluate the health status of the equipment, the weights of each parameter feature can flexibly adapt to changes in the operating status of the equipment, thereby improving the accuracy and timeliness of the health assessment.

[0047] Step S206, using a health assessment model to assess the target operating status data to obtain a device health score, wherein the device health score is used to quantitatively assess the health status of the device.

[0048] In the above step S206, the health assessment model includes a neural network model, a support vector machine (SVM), a decision tree and other models. The equipment health score is a specific value, and the output of the above model can be a percentage from 0 to 100, based on which the equipment health score is determined. The equipment health status is the performance and operating status of the equipment at a specific point in time, which is used to reflect the efficiency, reliability and potential failure risk of the equipment.

[0049] Since traditional health assessment models are based on simple statistical analysis or empirical rules, the robustness and generalization ability of the models are poor. In complex industrial environments, they cannot cope well with multivariate data and nonlinear relationships. In order to solve this problem, in some embodiments, a neural network model is used to calculate the health score: γ = f(W·X+b), where γ is the health score, W is the weight matrix, X is the input parameter matrix, b is the bias vector, and f is the activation function.

[0050] Through the above steps S202 to S206, by adjusting the initial weights of various parameter features in the collected operating status data, and then using the health assessment model to perform health assessment on the adjusted operating status data, the purpose of dynamically adjusting the weights of various parameter features according to actual working conditions during equipment operation is achieved, thereby achieving the technical effect of accurately reflecting the actual health status of the equipment, and further solving the technical problem that the related technology uses static weight factors to evaluate the health status of the equipment and cannot accurately reflect the actual health status of the equipment.

[0051] In order to improve the accuracy of the initial weights of parameter features, the historical operation data of the equipment can be trained to obtain the initial weights of the parameter features. Specifically, the initial weights corresponding to the various parameter features are determined in the following ways: historical sample data of the equipment is obtained, wherein the historical sample data includes: historical operation status data of the equipment and the real health score corresponding to the historical operation status data; the historical operation status data is predicted using a linear regression model to obtain a predicted health score; the loss function is used to calculate the loss value between the predicted health score and the real health score, and the parameters of the linear regression model are adjusted according to the loss value to obtain the initial weights corresponding to the various parameters.

[0052] In some embodiments of the present application, the linear regression model is:

[0053] γ=β0+β1χ1+β2χ2+……+β nχ n

[0054] Among them, γ is the health score of the device, χ i is the data of the i-th sensor, β i is the corresponding weight.

[0055] Loss functions include mean squared error (MSE), mean absolute error (MAE), etc. The initial weights obtained through historical data training can make the model more accurate in predicting equipment health and reduce misjudgment. It is suitable for equipment maintenance and fault prediction, effectively reducing maintenance costs and improving equipment availability and reliability.

[0056] During the operation of the equipment, the degree of influence of different parameter characteristics on the health status of the equipment will change dynamically. For example, in the early stage of equipment operation, vibration (an optional parameter characteristic) data may be considered the most important. At this time, the weight corresponding to the vibration data is greater than the weight corresponding to the temperature. As the equipment ages (for example, the aging index is greater than the preset threshold), the fluctuation of temperature (an optional parameter characteristic) may better reflect the health status of the equipment. At this time, the weight corresponding to the temperature is greater than the weight corresponding to the vibration data. Therefore, it is necessary to dynamically adjust the weights corresponding to various parameter characteristics as the equipment continues to run.

[0057] In an optional embodiment, the following steps can be used to adjust the initial weights corresponding to the various parameter features to obtain target operating status data: calculate a first health score corresponding to the various parameter features with the initial weights; use a preset model to predict the various parameter features with the initial weights to obtain a second health score corresponding to the various parameter features; calculate the error between the first health score and the second health score; adjust the initial weights corresponding to the various parameter features according to the error to obtain target weights corresponding to the various parameter features; determine the various parameter features with the target weights as the target operating status data.

[0058] The first health score is to use the currently collected multiple parameter features and initial weights, and obtain the health score of the current state of the device through a pre-set health score calculation formula, such as linear combination, weighted average, etc. This score is used to reflect the health status assessment of the device based on the current data and initial weights. The preset model used to calculate the second health score can be a machine learning model trained based on historical data, such as a neural network model, a support vector machine model, etc. The currently collected multiple parameter features are input into the preset model, and the model outputs the predicted device health status, that is, the second health score, which is based on the prediction ability of the model and is used to reflect the predicted value of the device health status under the current parameter features.

[0059] The error is the difference between the first health score and the second health score, which is used to reflect the deviation between the actual health score and the model-predicted health score under the current weight (initial weight) setting. The initial weight can be adjusted according to the error through a variety of algorithms, for example, machine learning algorithms (such as neural networks, support vector machines, etc.), that is, learning the importance of each parameter by training the model; gradient descent method, that is, adjusting the weight according to the gradient direction of the error, so that the health score predicted by the model is closer to the actual score (first health score); fuzzy logic algorithm, that is, using fuzzy sets and rules to dynamically adjust the weight according to the size of the error; genetic algorithm, that is, by simulating the natural evolution process, iteratively optimizing the weight so that the model adapts to the actual operating status of the equipment; empirical rules, that is, dynamically adjusting the weight through the rule engine based on expert knowledge and experience.

[0060] In order to facilitate understanding of the above process, an explanation is given below in conjunction with a specific embodiment:

[0061] For industrial pumps, their health status will be evaluated based on six quantitative parameter characteristics, including vibration, temperature, current, pressure, speed and sound. These sensor data are collected at each time point and given initial weights. To simplify the explanation, assume that the initial weights are: vibration: 0.2, temperature: 0.2, current: 0.2, pressure: 0.2, speed: 0.1, sound: 0.1.

[0062] Step 1: Calculate the first health score. Assume that the vibration value of the equipment at a certain moment is 1.5, the temperature is 60°C, the current is 15A, the pressure is 5bar, the speed is 1000rpm, and the sound is 70dB. Use the above initial weights to calculate the first health score. The calculation method of the first health score is as follows: First health score = 1.5*0.2+60*0.2+15*0.2+5*0.2+1000*0.1+70*0.1=80.

[0063] Step 2: Use the preset model to predict the second health score. Assume that the second health score predicted by the support vector machine model is: second health score=85.

[0064] Step 3: Calculate the error: Error = |first health score - second health score| = |80 - 85| = 5.

[0065] Step 4: Adjust the weights according to the error. If the error exceeds the preset threshold, it indicates that under the current weight distribution, the model's prediction results are significantly different from the actual score, and the weights need to be adjusted. When adjusting the weights, you can consider the following strategies: If the error of the vibration parameter is large, you can appropriately increase the weight of vibration and reduce the weights of other parameters in order to reduce the error of the next evaluation. The adjusted weights are: vibration: 0.25, temperature: 0.175, current: 0.175, pressure: 0.175, speed: 0.125, sound: 0.1.

[0066] Step 5: Determine the target operating status data. After adjusting the weights, a new weight distribution is obtained, namely the target weights. Various parameter features with target weights will be used to calculate subsequent health scores and become the target operating status data.

[0067] In order to ensure that the parameter characteristics can accurately reflect the health status of the equipment, multiple data of multiple parameters can be collected at a certain collection frequency and time interval to form a continuous data stream of the equipment operation status. By analyzing the data, the characteristic value of each parameter that best reflects the health status of the equipment is extracted. Specifically: the data of multiple parameters are collected at a preset collection frequency to obtain multiple corresponding parameter data sets collected at different preset time intervals for the multiple parameters; the parameter types of the multiple parameters are determined; for each parameter data set, the parameter type to which the parameter data in each parameter data set belongs is determined; the processing strategy corresponding to the parameter type is determined, wherein the processing strategy includes one of the following: extracting the maximum value, extracting the minimum value, calculating the average value, calculating the range, and calculating the variance; the multiple parameter data in each parameter data set are processed according to the processing strategy to obtain the parameter characteristics corresponding to each parameter data set in the multiple parameter characteristics.

[0068] The data of the above-mentioned multiple parameters are specific values ​​corresponding to the parameters. For example, multiple sensors are installed on an industrial pump to measure vibration, temperature and current respectively. By presetting the collection frequency during the operation of one day (i.e. 24 hours), the data recorded by the sensor are: vibration is 0.5-2.0mm / s (including multiple specific values ​​between 0.5-2.0), temperature is 50-80℃ (including multiple specific values ​​between 50-80), and current is 10-20A (including multiple specific values ​​between 10-20). The maximum value of the vibration parameter is extracted as the characteristic value, the maximum value of the temperature parameter is extracted as the characteristic value, and the average value of the current parameter is calculated as the characteristic value.

[0069] Related technologies usually judge the health status of equipment by regular evaluation or manual inspection. This method is not real-time and easily leads to delayed discovery of problems, and inability to perform maintenance or prevent failures in a timely manner. In order to solve this problem, after obtaining the actual health score of the equipment, maintenance information and / or warning information can also be generated according to the equipment health score. Specifically: the health score is compared with the preset maintenance threshold and the preset warning threshold, respectively, wherein the preset maintenance threshold is higher than the preset warning threshold; when the health score is lower than the preset maintenance threshold, maintenance suggestion information is generated; when the health score is lower than the preset warning threshold, warning information is generated. By continuously monitoring the health score of the equipment, an alarm can be triggered immediately and / or maintenance suggestions can be made when the health drops to the preset threshold, which can reduce unnecessary regular inspections and maintenance, reduce maintenance costs, and identify potential failures in advance and reduce the risk of unplanned downtime, thereby improving the efficiency and reliability of equipment management.

[0070] The initial weights corresponding to the various parameter features are adjusted, and before obtaining the target operating state data, the collected parameter features can also be preprocessed to ensure the accuracy and consistency of the data. For example, by denoising, the data can be smoothed and short-term fluctuations can be eliminated, so as to extract the main trend of the data and ensure the purity of the data; by normalization, the data of different dimensions can be normalized to the same range, so that the data of different parameters are comparable, which is convenient for subsequent weight calculation and health evaluation. Specifically: in response to the selection instruction, the target filtering method is determined from a plurality of types of filtering methods, wherein the plurality of types of filtering methods include median filtering and mean filtering; the target filtering method is used to remove the noise of the various parameter features, and the filtering data corresponding to the various parameter features are obtained; the minimum-maximum value normalization algorithm is used to normalize the filtering data, and the normalized data corresponding to the various parameter features are obtained. Among them, the above-mentioned selection instruction can be a selection instruction detected from a human-computer interaction interface, and the human-computer interaction interface is at least used to display a plurality of types of filtering methods. After the selection instruction is detected, the target filtering method is selected from the displayed plurality of types of filtering methods, and a filtering instruction is generated to filter the collected plurality of parameter features.

[0071] In some embodiments, median filtering can be used to denoise parameter features, and the calculation formula is:

[0072] filtered_data[i]=median(χ i-k , χ i-k+1 ,......,χ i+k )

[0073] Among them, filtered_data[i] is the value of the filtered signal at the i-th point, median is the median function, Xi-k , X i-k+1 , ..., X i+k is the value of 2k+1 data points around the i-th point in the original signal, and k is the window radius of the filter.

[0074] The minimum-maximum normalization calculation formula is:

[0075]

[0076] Among them, X i is the original data, min(X) is the minimum value in the data set, max(X) is the maximum value in the data set, X i ' is the normalized data.

[0077] Figure 3 is a schematic diagram of a device health assessment method according to an embodiment of the present application, such as Figure 3 As shown in Figure 1, this method can be divided into six stages: industrial field equipment data collection, raw data preprocessing, dynamic weight calculation, dynamic weight adjustment, health calculation, and evaluation result application (i.e. Figure 3 in stages 1-6).

[0078] The first stage: data collection, that is, collecting various data such as vibration, temperature, current, etc. of the equipment through multiple sensors installed at key positions of the equipment. The key position is the position where the parameter data that can best reflect the health status of the equipment can be collected. Through multi-dimensional data collection, the operating status of the equipment can be fully reflected, ensuring that the health assessment model can obtain comprehensive data input, providing sufficient data support for health assessment, and thus being able to more accurately reflect the health status of the equipment.

[0079] The second stage: data preprocessing, that is, further processing the collected parameter features by means of denoising, filtering, normalization, etc. Through the preprocessing step, the evaluation model can obtain cleaner and more unified data, ensuring the accuracy and consistency of the subsequent model input data, thereby improving the reliability and accuracy of the prediction results or evaluation results. After the data is preprocessed, the data can be stored to obtain historical data records, which can be used to calculate the initial weights.

[0080] The third stage: dynamic weight calculation, that is, using the dynamic weight algorithm to calculate the initial weight. The initial weight can be calculated based on the real-time working condition (referring to the actual operating status of the equipment) and historical data. For example, based on the actual working condition and historical data of the equipment, the initial dynamic weight of each measuring point parameter is calculated using a dynamic weight algorithm (such as a machine learning algorithm or a fuzzy logic algorithm).

[0081] The fourth stage: dynamic weight adjustment, that is, real-time adjustment of the initial weights corresponding to various parameter features. The initial weights can be adjusted according to real-time data and preset rules or models, so that the weights of parameter features can flexibly adapt to changes in the operating status of the equipment, and the evaluation model can reflect the status of the equipment in real time, thereby improving the accuracy and timeliness of health assessment and the accuracy and reliability of equipment health management.

[0082] The fifth stage: health calculation, that is, using the health assessment model to calculate the health score of the equipment. The pre-processed multiple parameter features and dynamically adjusted weights can be input into a comprehensive health assessment model (a health scoring algorithm model based on multi-dimensional parameter features and dynamic weights), such as neural networks, support vector machines and other models. By training and verifying multi-dimensional data through the model, the complex relationship between the equipment operating status data and health can be accurately captured, providing a more accurate and reliable health score. In an optional embodiment, the normalized vibration, temperature, current and sound data are input into the trained neural network model to obtain the health score of the motor, which can accurately reflect the current health status of the equipment.

[0083] The sixth stage: application of evaluation results, that is, judging the operating status of the equipment according to the health score, and providing corresponding maintenance suggestions and early warning information. In addition, the evaluation results can be displayed to the operator in real time through the display interface to help the operator understand the health status and potential risks of the equipment in a timely manner. For example, the early warning threshold is set according to the health score (for example, according to the value range of the health score). If the health score is lower than 0.3 (the health score drops to the preset threshold), the early warning is triggered to prompt the operator to inspect and maintain the equipment. The health score and early warning information are displayed in real time through the visual interface, and the health score of the motor is updated in real time on the display screen in the control room. When the score is lower than the threshold, an alarm pops up, thereby realizing real-time monitoring of the health status of the equipment, which can effectively avoid sudden failures of the equipment, reduce the risk of sudden failures and equipment downtime, and improve the efficiency and reliability of equipment management.

[0084] Figure 4 is a structural diagram of a device health assessment apparatus according to an embodiment of the present application, such as Figure 4 As shown, the device comprises:

[0085] The acquisition module 40 is used to acquire the operation status data of the device, wherein the operation status data includes: multiple parameter characteristics collected at multiple locations of the device;

[0086] An adjustment module 42 is used to adjust the initial weights corresponding to the various parameter features to obtain target operating state data;

[0087] The evaluation module 44 is used to evaluate the target operating status data using a health evaluation model to obtain a device health score, wherein the device health score is used to quantitatively evaluate the health status of the device.

[0088] In the acquisition module in the above-mentioned device for equipment health assessment, the acquisition module is also used to acquire historical sample data of the equipment, wherein the historical sample data includes: historical operating status data of the equipment and a real health score corresponding to the historical operating status data; a linear regression model is used to predict the historical operating status data to obtain a predicted health score; a loss function is used to calculate the loss value between the predicted health score and the real health score, and the parameters of the linear regression model are adjusted according to the loss value to obtain the initial weights corresponding to multiple parameters.

[0089] In the adjustment module in the above-mentioned device for equipment health assessment, the adjustment module is also used to calculate a first health score corresponding to multiple parameter characteristics with initial weights; use a preset model to predict multiple parameter characteristics with initial weights to obtain a second health score corresponding to the multiple parameter characteristics; calculate the error between the first health score and the second health score; adjust the initial weights corresponding to the multiple parameter characteristics according to the error to obtain target weights corresponding to the multiple parameter characteristics; determine the multiple parameter characteristics with target weights as target operating status data.

[0090] In the adjustment module in the above-mentioned device health assessment apparatus, the adjustment module is also used to calculate the target weight by the following formula: β i (t+1)=β i (t)+α(e(t)-e(t-1))·χ i Among them, β i (t) is the weight of the i-th sensor at time t, α is the frequency of weight adjustment, e(t) is the evaluation error at time t, X i is the data collected by the i-th sensor.

[0091] In the acquisition module in the above-mentioned equipment health assessment device, the acquisition module is also used to collect data of multiple parameters according to a preset collection frequency, and obtain corresponding multiple parameter data sets of the multiple parameters collected respectively at different preset time intervals; determine the parameter types of the multiple parameters; for each parameter data set, determine the parameter type to which the parameter data in each parameter data set belongs; determine the processing strategy corresponding to the parameter type, wherein the processing strategy includes one of the following: extracting the maximum value, extracting the minimum value, calculating the average value, calculating the range, and calculating the variance; process the multiple parameter data in each parameter data set according to the processing strategy, and obtain parameter features corresponding to each parameter data set among the multiple parameter features.

[0092] The above-mentioned device for equipment health assessment also includes a processing module 46, which is used to compare the health score with a preset maintenance threshold and a preset warning threshold, respectively, wherein the preset maintenance threshold is higher than the preset warning threshold; when the health score is lower than the preset maintenance threshold, maintenance suggestion information is generated; when the health score is lower than the preset warning threshold, warning information is generated.

[0093] In the processing module in the above-mentioned equipment health assessment device, the processing module is also used to respond to a selection instruction to determine a target filtering method from a plurality of types of filtering methods, wherein the plurality of types of filtering methods include median filtering and mean filtering; the target filtering method is used to remove noise of a plurality of parameter characteristics to obtain filtering data corresponding to the plurality of parameter characteristics; the minimum-maximum value normalization algorithm is used to normalize the filtering data to obtain normalized data corresponding to the plurality of parameter characteristics.

[0094] An embodiment of the present application also provides an electronic device, which includes a memory and a processor, wherein the memory is used to store program instructions; the processor is connected to the memory and is used to execute the steps of the method for device health assessment in each embodiment of the present application.

[0095] For example, the processor performs the following functions by executing program instructions stored in the memory:

[0096] Obtain the operating status data of the equipment, wherein the operating status data is a variety of parameter characteristics collected at multiple locations of the equipment; adjust the initial weights corresponding to the various parameter characteristics to obtain the target operating status data; use a health assessment model to evaluate the target operating status data to obtain the equipment health score, wherein the equipment health score is used to quantitatively evaluate the health status of the equipment.

[0097] An embodiment of the present application also provides a computer program product, including computer instructions, which, when executed by a processor, implement the steps of the method for device health assessment in each embodiment of the present application.

[0098] The embodiments of the present application also provide a computer program, which, when executed by a processor, implements the steps of the method for device health assessment in each embodiment of the present application.

[0099] The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0100] In the above embodiments of the present application, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.

[0101] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0102] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0103] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0104] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, disk or optical disk and other media that can store program codes.

[0105] The above is only a preferred implementation of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A method for equipment health assessment, characterized in that: include: Acquiring operation status data of the device, wherein the operation status data includes: multiple parameter characteristics collected at multiple locations of the device; Adjusting the initial weights corresponding to the various parameter features to obtain target operating state data; The target operating status data is evaluated using a health assessment model to obtain a device health score, wherein the device health score is used to quantitatively evaluate the health status of the device.

2. The method according to claim 1, characterized in that The initial weights corresponding to the various parameter features are determined in the following manner: Acquire historical sample data of the device, wherein the historical sample data includes: historical operating status data of the device and a real health score corresponding to the historical operating status data; Using a linear regression model to predict the historical operating status data to obtain a predicted health score; A loss function is used to calculate a loss value between the predicted health score and the actual health score, and the parameters of the linear regression model are adjusted according to the loss value to obtain initial weights corresponding to the various parameters.

3. The method according to claim 1, characterized in that The initial weights corresponding to the various parameter features are adjusted to obtain target operating state data, including: Calculating a first health score corresponding to the plurality of parameter features having the initial weights; Using a preset model to predict the multiple parameter features with the initial weights to obtain second health scores corresponding to the multiple parameter features; calculating an error between the first health score and the second health score; Adjusting the initial weights corresponding to the various parameter features respectively according to the errors to obtain target weights corresponding to the various parameter features respectively; The plurality of parameter features having the target weights are determined as target operating state data.

4. The method according to claim 3, characterized in that The initial weights corresponding to the various parameter features are adjusted according to the errors to obtain target weights corresponding to the various parameter features, including: The target weight is calculated by the following formula: b i (t+1)=β i (t)+α(e(t)-e(t-1))·x i Among them, β i (t) is the weight of the i-th sensor at time t, α is the frequency of weight adjustment, e(t) is the evaluation error at time t, X i The data collected by the i-th sensor.

5. The method according to claim 1, characterized in that The various parameter characteristics are determined in the following manner: Collect data of the multiple parameters at a preset collection frequency to obtain corresponding multiple parameter data sets of the multiple parameters collected at different preset time intervals; determining parameter types of the plurality of parameters; For each parameter data set, determining the parameter type to which the parameter data in each parameter data set belongs; Determine a processing strategy corresponding to the parameter type, wherein the processing strategy includes one of the following: extracting a maximum value, extracting a minimum value, calculating an average value, calculating a range, and calculating a variance; The multiple parameter data in each parameter data set are processed according to the processing strategy to obtain parameter features corresponding to each parameter data set among the multiple parameter features.

6. The method according to claim 5, characterized in that The method further comprises: comparing the health score with a preset maintenance threshold and a preset warning threshold respectively, wherein the preset maintenance threshold is higher than the preset warning threshold; When the health score is lower than a preset maintenance threshold, generating maintenance suggestion information; When the health score is lower than a preset warning threshold, a warning message is generated.

7. The method according to any one of claims 1 to 6, characterized in that: Before adjusting the initial weights corresponding to the plurality of parameter features to obtain the target operating state data, the method further includes: In response to the selection instruction, determining a target filtering mode from a plurality of types of filtering modes, wherein the plurality of types of filtering modes include median filtering and mean filtering; The target filtering method is used to remove noise of the multiple parameter features to obtain filtering data corresponding to the multiple parameter features respectively; The filtered data is normalized using a minimum-maximum normalization algorithm to obtain normalized data corresponding to the multiple parameter features.

8. A device for equipment health assessment, characterized in that: include: An acquisition module, used to acquire the running status data of the device, wherein the running status data is a plurality of parameter characteristics collected at a plurality of positions of the device; An adjustment module, used to adjust the initial weights corresponding to the various parameter features to obtain target operating state data; The evaluation module is used to evaluate the target operating status data using a health evaluation model to obtain a device health score, wherein the device health score is used to quantitatively evaluate the health status of the device.

9. An electronic device, characterized in that: include: A memory and a processor, wherein the memory is used to store program instructions; the processor is connected to the memory and is used to execute the method for device health assessment as described in any one of claims 1 to 7.

10. A computer program product comprising computer instructions, characterized in that When the computer instructions are executed by a processor, the method for device health assessment described in any one of claims 1 to 7 is implemented.

Images