Controlled operating component failure monitoring processing logic method based on multi-source operating data fusion

Abstract

This invention relates to the fields of monitoring and control and data acquisition technology, and discloses a fault monitoring and processing logic method for controlled operating components based on multi-source operational data fusion. The method includes: acquiring the drive-end current command sequence and the physical feedback parameter sequence; calculating the rise rate and determining the response start time when the command sequence undergoes a step change; extracting command execution hysteresis characteristics and calculating cross-correlation coefficients to calibrate the phase offset value; correcting the sampling window phase; calculating the state residual sequence; and outputting a fault processing command when the sequence exceeds a preset monitoring threshold. This invention establishes a phase self-anchoring mechanism based on the correlation of command excitation wavefronts, eliminating dynamic response deviations caused by physical entity attribute drift, and achieving adaptive decoupling between monitoring sensitivity and load fluctuations. It is particularly suitable for vertical lifting systems with long-stroke flexible traction components, improving the accuracy of identifying sub-health states of controlled components by compensating for timing jitter caused by the release of elastic potential energy of the traction medium.

Description

Technical Field

[0001] This invention belongs to the field of monitoring and control and data acquisition technology, and in particular relates to a logic method for monitoring and processing faults of controlled operating components based on the fusion of multi-source operating data. Background Technology

[0002] Current monitoring, control, and data acquisition systems are core architectures for ensuring the safe operation of controlled components. They collect operating parameters such as current, speed, or position of the controlled components and compare them with pre-set static threshold ranges to determine the current operating status of the controlled components. There is an inherent physical response lag between the physical response of the controlled components and the issuance of control commands, caused by system dynamics inertia, communication link delays, and the physical characteristics of the actuators. During the transient evolution phase of the controlled process, there is a time-series misalignment between the energy injection wavefront at the drive end and the power consumption response wavefront at the actuator end. Existing monitoring logic tends to treat real-time feedback data as a static mirror image that occurs synchronously with control commands. This logical assumption ignores the time-series offset in the dynamic process, resulting in logical residuals under high-frequency variable operating conditions such as acceleration, deceleration, or reversal of motion polarity.

[0003] To avoid frequent alarms caused by physical lag, methods such as widening the judgment threshold or increasing the signal filtering depth are adopted. However, simply increasing the threshold width can mask the slight deviations of controlled operating components in the early stages of sub-health, leading to the inability to effectively detect early deterioration phenomena such as increased mechanical friction or localized temperature rise in windings. Simple signal smoothing introduces additional logical lag, further reducing the system's real-time response capability to sudden load fluctuations or feedback link interference. Although the industry has attempted to compensate for lag by improving sensor accuracy or introducing model prediction algorithms, such methods often increase additional hardware sensing costs or are difficult to deploy in general industrial scenarios due to excessive computational overhead. In addition to the hard deviations caused by the physical properties of the actuator, the existing monitoring logic processes electromagnetic parameters and mechanical feedback with poor correlation. At the same time, it also shows insufficient adaptability to dynamic operating conditions. For example, Chinese invention patent with authorization announcement number CN101680457B discloses a device and method for fault monitoring. It transforms the centrifugal pump head characteristic line point to the motor parameter diagram, constructs a preset allowable operating area, and determines whether the motor operating point exceeds the limit. Although it considers the parameter correspondence under variable speed conditions, it is essentially still a quasi-static spatial threshold monitoring. When the controlled component performs high-frequency transient actions, this logic does not touch the time domain misalignment between the command excitation wavefront and the physical response wavefront. It cannot calibrate and eliminate the real-time phase drift caused by the system dynamic inertia. Under variable load conditions, the monitoring system is difficult to accurately decouple physical hysteresis fluctuations and actual functional degradation. It needs to make a compromise between sacrificing alarm sensitivity and maintaining operational stability.

[0004] Therefore, how to achieve dynamic alignment of multi-source operating data based on the changing characteristics of control commands, and effectively distinguish between physical lag and functional abnormality, is the technical problem to be solved by this invention. Summary of the Invention

[0005] This invention provides a fault monitoring and processing logic method for controlled operating components based on multi-source operational data fusion, comprising the following steps: Step S1: Obtain the drive end current command sequence of the controlled operating component and the feedback parameter sequence characterizing the physical response state of the controlled operating component; Step S2: When a step change occurs in the driving end current command sequence, calculate the step rise rate of change of the driving end current command sequence, and determine the response start time corresponding to the step rise rate of change based on the sampled values ​​of the feedback parameter sequence. Step S3: Calculate the time deviation between the step rise rate and the response start time using time-domain difference, and define the time deviation as a dynamic deviation characteristic value that characterizes the instruction execution lag. Step S4: Using the dynamic deviation characteristic value as the search center, calculate the cross-correlation coefficient between the drive end current command sequence and the feedback parameter sequence within the sliding window period, and calibrate the time delay when the cross-correlation coefficient reaches its maximum value as the real-time phase offset value. Step S5: Use the real-time phase offset value to perform phase synchronization compensation correction on the preset monitoring sampling window to generate a time-aligned monitoring data stream. Step S6: Input the monitoring data stream into the preset deviation evaluation logic. The deviation evaluation logic is configured to calculate the difference between the monitoring data stream and the reference data stream to generate the operating state residual sequence, and output the fault handling instruction for the controlled operating component when the amplitude of the operating state residual sequence continuously exceeds the preset monitoring threshold.

[0006] Preferably, step S6 includes: acquiring real-time load data of the controlled operating component, and performing linear scaling correction on the operating state residual sequence based on the real-time load data to offset the amplitude shift of the operating state residual sequence caused by load fluctuations.

[0007] Preferably, step S1 further includes: calculating the variance of the rate of change of the feedback signal and establishing a signal quality assessment mechanism accordingly; identifying and eliminating random electromagnetic pulse interference generated by the controlled environment by mapping the change characteristics of the feedback parameter sequence with the logic state of the drive end current command sequence; and when it is determined that a non-physical step change occurs in the feedback parameter sequence, using the operating state residual sequence of the previous sampling period to perform state prediction.

[0008] Preferably, after step S4, the method further includes: extracting trend terms that exhibit a monotonic change trend from the residual sequence of the operating state, converting the trend terms into update step sizes of the time constants of the controlled operating components, and performing online parameter calibration on the preset monitoring reference model.

[0009] Preferably, after step S5, the method further includes: superimposing a high-frequency micro-excitation disturbance signal onto the current command sequence at the drive end during the steady-state operation cycle, and obtaining the frequency response characteristics of the feedback parameter sequence to the high-frequency micro-excitation disturbance signal; and extracting structural response indices at a specific frequency. Monitor the structural connection status and structural response indicators of the controlled operating components. The calculation method is as follows: ,in, In frequency The measured impedance amplitude, in units of , This represents the phase response offset angle at the corresponding frequency, in rad. This represents the total number of sampling frequency points.

[0010] Preferably, in step S1, the feedback parameter sequence includes 50Hz data representing the rotation frequency, 10mm data representing the displacement, and 0.5mm data representing the temperature rise rate. / s data; multi-source runtime data is mapped to a multi-dimensional monitoring space consisting of time, logical state, and execution intensity through a normalization operator.

[0011] Preferably, in step S3, the dynamic deviation characteristic value is obtained by calculating the time-domain overlap integral between the step rise rate and the response start time, and the calculation step size of the real-time phase offset value is dynamically corrected by using the amplitude of the time-domain overlap integral.

[0012] Preferably, in step S6, the preset monitoring threshold is calibrated based on the health benchmark database of the controlled operating component; when the operating status residual sequence shows a monotonically increasing trend within a time period of 500ms to 2000ms and the slope exceeds the preset rate of change threshold, the controlled operating component is determined to be in a state of performance degradation.

[0013] Preferably, the fault handling instructions include a shutdown instruction for driving the controlled operating component into a protection mode, a fault alarm signal for display on the terminal, and maintenance work order data containing fault characteristic quantities.

[0014] Preferably, the fault monitoring and processing logic method is deployed between the field controller and the decision server to execute instruction compensation when the controlled operating component experiences an operational deviation.

[0015] Compared with existing technologies, the controlled operating component fault monitoring and processing logic method based on multi-source operating data fusion of the present invention has the following advantages: 1. In the fault monitoring and processing logic of the operating components, the control command sequence change rate is used to map the logic phase offset, and the timing normalization processing of multi-source operating data is performed in conjunction with the control system to solve the physical response hysteresis interference caused by the dynamic characteristics of the actuator. By identifying the time difference between the energy injection moment of the drive end and the power consumption response moment of the execution end at the moment of command start, the heterogeneous data streams are aligned in near real-time at the logic level, eliminating the pseudo residuals generated during the switching of operating conditions, and enabling the monitoring logic to accurately distinguish between normal physical hysteresis fluctuations and substantial functional abnormalities.

[0016] 2. By extracting the first and second derivative features of the driving parameters, a load response operator is generated, and nonlinear scaling correction is performed on the logic phase offset accordingly. This achieves adaptive decoupling between monitoring sensitivity and variable load conditions. The dynamic characteristics of energy input are used to reversely deconstruct physical load changes. Without altering the hardware sensing link, the phase sliding window inaccuracy caused by environmental parameter drift or load fluctuation is offset, thereby improving the accuracy of identifying the sub-health state of components under variable load conditions.

[0017] 3. Introduce a feedback link anomaly isolation mechanism based on local information entropy of response parameters. By synchronously mapping data change characteristics with the logical state grid of the instruction domain, it identifies and shields random pulse interference generated by the industrial electromagnetic environment. When a non-physical random change is detected in the feedback signal, the data suspension logic is activated and virtual prediction is performed using historical residual trends. This maintains the logical continuity of the monitoring process and prevents signal source pollution from inducing unexpected fault protection command output. Attached Figure Description

[0018] Fig. 1 This is a flowchart of the controlled operating component fault monitoring process based on phase self-anchoring and timing alignment of the present invention. Fig. 2 This is a closed-loop monitoring logic architecture diagram of the present invention, which integrates adaptive load correction and active structural auditing. Detailed Implementation

[0019] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0020] It should be noted that all directional and positional terms used in this invention, such as: up, down, left, right, front, back, vertical, horizontal, inner, outer, top, low, lateral, longitudinal, center, etc., are only used to explain the relative positional relationship and connection between components in a specific state (as shown in the accompanying drawings). They are only for the convenience of describing this invention and do not require that this invention be constructed and operated in a specific orientation. Therefore, they should not be construed as limiting this invention. In addition, the descriptions of "first," "second," etc., in this invention are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly indicating the number of technical features indicated.

[0021] In the description of this invention, unless otherwise explicitly specified and limited, the terms installation, connection, and linking should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to mechanical connections; they can refer to direct connections or indirect connections through an intermediate medium; they can refer to the internal communication between two components. For those skilled in the art, the specific meaning of the above terms in this invention can be understood according to the specific circumstances.

[0022] In the description of this specification, references to the terms "an embodiment," "some embodiments," "illustrative embodiments," "examples," "specific examples," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example, and the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0023] This invention provides a fault monitoring and processing logic method for controlled operating components based on multi-source operational data fusion. The method uses the acquired drive-end current command sequence and physical feedback parameter sequence as logical inputs. By identifying the excitation wavefront correlation during command step changes, a phase self-anchoring mechanism is established. The system calculates the rise rate and determines the response start time, extracts command execution hysteresis features, and calibrates the real-time phase offset value, thereby performing phase synchronization compensation on the monitoring sampling window. Based on this, the system calculates the time-aligned state residual sequence and outputs a fault processing command when it exceeds a preset monitoring threshold. This method achieves adaptive decoupling of monitoring sensitivity and load fluctuations through logical alignment of energy characteristics, improving the identification accuracy of controlled components in the early stages of performance degradation. In the specific monitoring execution process, the system executes the following steps... The system acquires the drive-end current command sequence of the controlled operating component and the feedback parameter sequence characterizing the physical response state of the controlled operating component; the feedback parameter sequence includes a sequence characterizing the rotation frequency. Data, characterizing displacement Data and characterization of temperature rise rate Data; To ensure the quality of the feedback signal, the system establishes a signal quality assessment mechanism by calculating the variance of the feedback signal change rate, and synchronously maps the change characteristics of the feedback parameter sequence with the logic state execution of the drive end current command sequence to identify and eliminate random electromagnetic pulse interference generated by the controlled environment; When it is determined that a non-physical step change occurs in the feedback parameter sequence, the system uses the operating state residual sequence of the previous sampling period to perform state prediction to maintain the continuity of the monitoring logic.

[0024] Execution steps When a step change occurs in the drive current command sequence, the rate of change of the step rise edge of the drive current command sequence is calculated, and the response start time corresponding to the rate of change of the step rise edge is determined based on the sampled values ​​of the feedback parameter sequence; then the following steps are executed. The time deviation between the step rise rate and the response start time is calculated using time-domain difference, and this time deviation is defined as a dynamic deviation characteristic value characterizing the instruction execution lag. This dynamic deviation characteristic value is obtained by calculating the time-domain overlap integral between the step rise rate and the response start time. When determining the response start time, the feedback parameter sequence is calculated in... Standard deviation within the sliding window In the sampling sequence after a step change in the drive current command sequence, the rate of change of the feedback value is continuously... More than one sampling period The first sampling point was identified as the start time of the physical response, and structural response indices were obtained. impedance amplitude At that time, a Fast Fourier Transform is performed on the drive current command sequence and the feedback parameter sequence to calculate their frequency response. The ratio of the spectral modulus values ​​converts the current command energy input and the physical entity displacement feedback into a unified frequency domain transfer function characteristic quantity, eliminating quantization ambiguity when heterogeneous data sources are directly compared in the time domain; after obtaining the dynamic deviation characteristic value, the system executes the following steps. Using the dynamic deviation characteristic value as the search center, the cross-correlation coefficient between the drive end current command sequence and the feedback parameter sequence within the sliding window period is calculated, and the time delay when the cross-correlation coefficient reaches its maximum value is calibrated as the real-time phase offset value. During this process, the system uses the amplitude of the time-domain overlap integral to dynamically correct the calculation step size of the real-time phase offset value. In addition, the trend term that shows a monotonically changing trend in the operating state residual sequence is extracted, and the trend term is converted into the update step size of the time constant of the controlled operating component, and online parameter calibration is performed on the preset monitoring reference model.

[0025] Next steps The system uses real-time phase offset values ​​to perform phase synchronization compensation correction on a preset monitoring sampling window, generating a time-aligned monitoring data stream. After this step, a high-frequency micro-excitation disturbance signal is superimposed on the current command sequence at the drive end during the steady-state operation cycle, and the frequency response characteristics of the feedback parameter sequence to this high-frequency micro-excitation disturbance signal are obtained. The system then extracts structural response indices at specific frequencies. To monitor the structural connection status and structural response indicators of controlled operating components. The calculation formula is as follows: ,in, For structural response indicators; In frequency The measured impedance amplitude, in units of ; The phase response offset angle at the corresponding frequency, in units of ; This represents the total number of sampling frequency points; the final step is to execute the following steps. The monitoring data stream is input into a preset deviation evaluation logic to calculate the difference between the monitoring data stream and the baseline data stream to generate an operating state residual sequence. When the amplitude of this operating state residual sequence continuously exceeds a preset monitoring threshold, a fault handling instruction for the controlled operating component is output. The preset monitoring threshold is calibrated based on the health baseline database of the controlled operating component. When the operating state residual sequence... to When the controlled operating component exhibits a monotonically increasing trend within a time period and the slope exceeds a preset rate of change threshold, it is determined that the controlled operating component is in a state of performance degradation. Furthermore, the system acquires real-time load data of the controlled operating component and performs linear scaling correction on the operating state residual sequence based on this real-time load data to offset amplitude shifts caused by load fluctuations. The scaling correction of the operating state residual sequence and the monitoring threshold are determined through an offline calibration process to obtain compensation gains. Before the controlled operating component is put into operation, the compensation gains are determined at rated loads. , and Execute standard step commands at quantile points, record the mean residual value corresponding to the load gradient, and fit the load compensation coefficient using the least squares method. Collect data on the continuous operation of controlled components under health baseline conditions. The energy coupling residuals for each sampling period are calculated as the arithmetic mean. with standard deviation Set the preset monitoring threshold Determined as This allows for dynamic compensation of residual amplitude based on real-time load data during online monitoring, ensuring that residual fluctuations are anchored within a linear range determined by the real-time load of the physical entity. The final output fault handling instructions include shutdown instructions to drive controlled operating components into protection mode, fault alarm signals to be displayed on the terminal, and maintenance work order data containing fault characteristic quantities. This monitoring and processing logic method is deployed between the field controller and the decision server to execute instruction compensation when component operating deviations occur.

[0026] Example 1: When the controlled operating component performs a vertical stroke lifting action, the drive end current command sequence is in Internal occurrence to The step change, and the combined load on the suspension side switches between rated quantile points as the occupants enter and exit; under this condition, due to the dynamic lag of the long-stroke flexible traction component, the feedback parameter sequence of the actuator and the current command sequence of the drive end will produce timing discrepancies. The phase misalignment causes the calculated energy coupling residual sequence to generate amplitude components exceeding the monitoring threshold, triggering a non-fault-related abnormal alarm signal. The system then executes the monitoring and processing logic of the specific implementation method, and the rate of change reaches [a certain value] at the step rising edge. At that time, the phase self-anchoring mechanism based on the correlation of the excitation wavefront is used to lock the energy injection time. With power consumption response time The real-time phase offset value is calibrated by calculating the cross-correlation coefficient between the drive current command sequence and the feedback parameter sequence within the sliding window period. The system uses the real-time phase offset value to compensate the sampling window, ensuring the aligned input power... With effective power By achieving feature overlap, the operating condition coupling component in the energy coupling residual sequence is reduced to the level of background noise.

[0027] During the steady-state operating cycle, the system adds an amplitude equal to the rated command amplitude to the current command sequence sent to the drive end. The system obtains the frequency response characteristics of the feedback parameter sequence from the high-frequency micro-excitation disturbance signal; the system calculates the structural response index based on the frequency response characteristics. The structural response index The calculation formula is as follows: ,in, For structural response indicators; In frequency The measured impedance amplitude, in units of ; The phase response offset angle at the corresponding frequency, in units of ; The total number of sampling frequency points; when the connection of the controlled operating component becomes loose, the phase response offset angle. Increased, leading to structural response indicators It exhibits a monotonically increasing trend. Since the real-time phase offset value eliminates the physical response hysteresis interference caused by frequent load switching, the system maintains the identification of component performance degradation under variable load conditions. Based on the trend evolution of the residual sequence of operating status, it determines that the controlled operating component is in a sub-healthy state and outputs fault handling instructions containing shutdown instructions and maintenance work order data. Under the residual monitoring after phase alignment, the controlled operating component realizes the early warning output of the initial stage of physical degradation.

[0028] Example 2: The test verification was initiated in the physical simulation environment of the industrial servo actuator monitoring platform. The data of the test platform was acquired through a high-frequency data acquisition unit, which has the following functions: Sampling frequency and Bit-to-digital conversion resolution; sampling period The setting balances the acquisition accuracy of wavefront features with the computational overhead of the control system, when the spectral bandwidth of the monitored signal reaches... At that time, set for To satisfy the sampling theorem and prevent signal aliasing; the superimposed signal-to-noise ratio in the experimental signal source is... Gaussian white noise and frequency of The power frequency interference harmonics are used to simulate the electromagnetic environment characteristics of industrial sites. The data originates from a physical experimental platform, which synchronously acquires drive energy flow and work response data through a high-precision current sensor and a displacement encoder.

[0029] To verify the effectiveness of the present invention in eliminating false alarms caused by response lag, a control group using a preset static inertia constant compensation logic and an experimental group using the excitation wavefront self-anchoring mechanism of the present invention were set up; the current command sequence at the driving end was executed. to During the step change process, the energy injection time is locked. With power consumption response time Real-time calibration of real-time phase offset values; Table 1 is a comparison table of recorded state monitoring data under different load gradients, see Table 1; the original input parameters include load quality, the key intermediate feature value is the aligned phase offset, and the final output is the residual of the determined operating state; the measured values ​​reflect the change with load quality from... Increase to The residual amplitude of the control group was changed due to time misalignment. linear growth to The experimental group, after phase synchronization compensation correction, maintained its energy coupling residual amplitude at [value missing]. to Within the interval; by calculating the variance of the rate of change of the feedback signal, the system identifies and eliminates superimposed [variables / indicators]. Random impulse interference maintains the logical continuity of the residual sequence.

[0030] Table 1: Comparison of Status Monitoring Data Recorded Under Different Load Gradients

[0031] When the load quality is improved to At that time, due to the actuator entering the electromagnetic saturation region and the physical inertia exceeding the effective adjustment step size of the system logic phase offset, the measured operating state residuals exhibit a nonlinear surge trend; the phase response offset angle The measured value is in Under operating conditions higher than Under working conditions ; confirm the limitations to The load range is the preferred operating window to ensure the closed loop of energy coupling residual logic; through structural response indicators Through continuous monitoring, the system identifies that Small trend term shifts within the range, when The index is from steady state Linear evolution to At that time, the system detected a decrease in the stiffness of the connector. Physical signs.

[0032] Example 3: This example combines Figs. 1-2 The following describes the logic and method for fault monitoring and processing of controlled operating components based on multi-source operational data fusion, such as... Fig. 1As shown, the monitoring and processing logic method executes step S1 to obtain the drive-end current command sequence of the controlled operating component and the feedback parameter sequence characterizing the physical response state of the controlled operating component. Step S2 calculates the step rise rate when the drive-end current command sequence undergoes a step change, and determines the response start time corresponding to the step rise rate based on the sampled values ​​of the feedback parameter sequence. Then, step S3 calculates the time deviation between the step rise rate and the response start time using time-domain calculus. The time deviation is defined as a dynamic deviation characteristic value characterizing the command execution lag. Step S4 is executed, using the dynamic deviation characteristic value as the search center to calculate the cross-correlation coefficient between the drive end current command sequence and the feedback parameter sequence within the sliding window period, and calibrating the time delay when it reaches its maximum value as the real-time phase offset value. Then, step S5 is executed to use the real-time phase offset value to perform phase synchronization compensation correction on the preset monitoring sampling window, generating a timing-aligned monitoring data stream. Finally, step S6 is executed to input the monitoring data stream into the preset deviation evaluation logic to calculate the operating state residual sequence, and output a fault handling command when the amplitude of the operating state residual sequence continuously exceeds the preset monitoring threshold.

[0033] like Fig. 2 As shown, the logical architecture constructs a closed-loop monitoring network between the field controller, decision server, and operation and maintenance display terminal. The field controller is responsible for collecting input commands and feedback data, injecting micro-excitation signals, performing phase self-anchoring to establish the correlation of command excitation wavefronts, and entering the execution fault monitoring and evaluation stage. When the residual is detected to be greater than the threshold, the fault handling command module is triggered, which includes shutdown, alarm, and work order information and is transmitted to the decision server and operation and maintenance display terminal. The load adaptive correction module performs linear scaling to offset the impact of load fluctuations. At the same time, the structural health active audit module and the model parameter online calibration module are connected in parallel at the output end. The former analyzes the structural status through high-frequency micro-excitation response, and the latter updates the model parameters online through trend term extraction, thus forming a multi-dimensional controlled operating component fault monitoring and processing logical system.

[0034] Example 4: During the long-cycle operation of a high-power vertical traction drive, the response characteristics between the vibration feedback of the traction wheel support and the drive current command of the controlled operating component change with the ambient temperature of the drive room and the operating temperature of the friction pair. Rise to The monotonic shift causes a deviation in the preset monitoring reference model. To determine the numerical boundaries of the monitoring threshold, the system acquires data on the controlled operating components after the pump unit enters a stable operating state. Calculate the arithmetic mean of the energy-coupled residual data points within each sampling period. with standard deviation and preset monitoring thresholds Determined as ;in, This represents the arithmetic mean of the energy coupling residuals; The standard deviation of the energy coupling residual; To preset monitoring thresholds, fault judgment boundaries are determined using statistical distribution characteristics. Cross-correlation coefficients are calculated using discrete convolution operations to calibrate real-time phase offset values. Specifically, a length of [missing information] is extracted from memory. instruction vector With feedback vector And execute the sliding window cross-correlation algorithm to calculate the time delay offset. The relevant components below Its calculation formula is ;in, The time delay offset is Cross-correlation components at time; For the instruction vector One element; The feedback vector is the first One element; For sampling index; The length of the vector; For the time delay offset, the system searches for... Index value that reaches the maximum value and combined with the sampling period Determine the real-time phase offset value ;in, This is the real-time phase offset value. This is the index value corresponding to the maximum value of the cross-correlation component. The sampling period.

[0035] To address parameter deviations caused by ambient temperature drift, the system executes a parameter correction procedure and continuously analyzes the residual sequence of operating status. Mean shift component within And according to the formula The component's inertial time constant is updated online; among which, The updated component inertial time constant; This is the current component inertial time constant; This is the mean shift component; The adaptive gain coefficient is set to . The system regenerates the monitoring data stream using the calibrated component inertial time constant, causing the centroid of the energy coupling residual distribution to return to within the preset monitoring threshold, thus maintaining the stability of the monitoring system under full temperature range conditions. After the pump unit reaches steady-state operation, the system superimposes the current command sequence to the drive end at a frequency of [frequency missing]. And the amplitude is The high-frequency micro-excitation perturbation signal is used to obtain the frequency response characteristics of the feedback parameter sequence to this signal, and the structural response index is calculated. The formula for quantifying mechanical strength loss is as follows: ,in, For structural response indicators; In frequency The measured impedance amplitude, in units of ; The phase response offset angle at the corresponding frequency, in units of ; The total number of sampling frequency points, when the system monitors structural response indicators. The calculated value is from Increase to When it is determined that the connection stiffness of the controlled operating component is weakened due to loose anchor bolts, a maintenance work order containing the position coordinates is output simultaneously.

[0036] Example 5: During the on-site deployment of a high-power industrial feedwater pump set, the system initiates an initial health benchmark calibration procedure for the on-site environment to obtain the controlled operating components under no-load steady-state conditions. Calculate the feedback parameter sequence from 10 consecutive sampled data points. Average amplitude within the sliding window and the dynamic variance of the sliding window ,in, This represents the average amplitude of the sliding window, in physical units of the feedback parameter. For the dynamic variance of the sliding window, the system identifies that the values ​​in the feedback parameter sequence exceed... The discrete sampling points are then eliminated.

[0037] The system injects a current command sequence with an amplitude equal to the rated command amplitude into the drive terminal. The step signal execution parameter sensitivity optimization is performed, and the convergence time of the residual sequence of the operating state returning to the preset monitoring threshold range is monitored in real time. The adaptive gain coefficient is determined according to the ratio of the convergence time to the preset decay time constant. The value of, among which, This is a dimensionless parameter update step size adjustment factor. When the system detects that the parameter update step size is continuously... The fluctuation amplitude within each sampling period remains at When the value is within the range, the adaptive calibration logic of the controlled operating component is adapted to the current physical damping characteristics of the field, so that the amplitude variance of the operating state residual sequence is kept in the healthy range, and the physical reference deviation caused by the difference in deployment environment is offset.

[0038] Example 6: In a scenario where a supervisory control and data acquisition system with multiple vertical lifting actuators is deployed, considering the individual differences in the length of the traction medium and the balance compensation mechanism, the system executes a standardized health baseline adaptive initialization procedure upon initial startup. This controls the controlled operating components to perform unloaded leveling inspection within a preset stroke. Simultaneously, a continuous sampling mode for the feedback parameter sequence is activated, and the distribution trajectory of the energy coupling residual is recorded. A residual statistical distribution model is constructed using the cleaned data points, and the arithmetic mean of the data is calculated. with standard deviation To determine the background logic noise in the current environment, the calculated value is written as a reference coefficient into the logic layer's storage unit, setting a preset monitoring threshold. Dynamically associate with the current state of physical entities; whereby, The arithmetic mean of the energy coupling residuals. The standard deviation of the energy coupling residual. This is the preset monitoring threshold.

[0039] When the system faces communication link jitter or heterogeneous protocol packet forwarding delays, the system determines the length parameter of the sliding window period by executing a time delay sensitivity calibration process and monitors the characteristic period of the drive end current command sequence in real time. With the fundamental period of the feedback parameter sequence The system determines the correlation search range based on the highest-order component of the physical motion frequency, and calculates the autocorrelation coefficient convergence performance under different window lengths. The system then sets the sliding window period to the period of the fundamental frequency of the feedback parameter sequence. to To capture the physical hysteresis envelope and eliminate phase window misalignment, this procedure correlates the reconstruction logic of the discrete sampling sequence with the variation characteristics of the physical execution frequency, thereby enabling real-time phase offset values ​​to be multiplied. The calculation step size is within the working range determined by the physical bandwidth constraint.

[0040] The embodiments of this application have been described above with reference to the accompanying drawings. Unless otherwise specified, the embodiments and features in the embodiments of this application can be combined with each other. This application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many modifications and improvements under the guidance of this application without departing from the spirit and scope of the claims. All such modifications and improvements are within the scope of protection of this application.

Claims

1. A fault monitoring and processing logic method for controlled operating components based on multi-source operational data fusion, characterized in that, Includes the following steps: Step S1: Obtain the drive end current command sequence of the controlled operating component and the feedback parameter sequence characterizing the physical response state of the controlled operating component; Step S2: When a step change occurs in the driving end current command sequence, calculate the step rise rate of change of the driving end current command sequence, and determine the response start time corresponding to the step rise rate of change based on the sampled values ​​of the feedback parameter sequence. Step S3: Calculate the time deviation between the step rise rate and the response start time using time-domain difference, and define the time deviation as a dynamic deviation characteristic value that characterizes the instruction execution lag. Step S4: Using the dynamic deviation characteristic value as the search center, calculate the cross-correlation coefficient between the drive end current command sequence and the feedback parameter sequence within the sliding window period, and calibrate the time delay when the cross-correlation coefficient reaches its maximum value as the real-time phase offset value. Step S5: Use the real-time phase offset value to perform phase synchronization compensation correction on the preset monitoring sampling window to generate a time-aligned monitoring data stream. Step S6: Input the monitoring data stream into the preset deviation evaluation logic. The deviation evaluation logic is configured to calculate the difference between the monitoring data stream and the reference data stream to generate the operating state residual sequence, and output the fault handling instruction for the controlled operating component when the amplitude of the operating state residual sequence continuously exceeds the preset monitoring threshold.

2. The controlled operating component fault monitoring and processing logic method based on multi-source operating data fusion according to claim 1, characterized in that, Step S6 includes: acquiring real-time load data of the controlled operating component, and performing linear scaling correction on the operating state residual sequence based on the real-time load data to offset the amplitude shift of the operating state residual sequence caused by load fluctuations.

3. The controlled operating component fault monitoring and processing logic method based on multi-source operating data fusion according to claim 1, characterized in that, Step S1 also includes: calculating the variance of the rate of change of the feedback signal and establishing a signal quality assessment mechanism accordingly; identifying and eliminating random electromagnetic pulse interference generated by the controlled environment by mapping the change characteristics of the feedback parameter sequence with the logic state of the drive current command sequence; and using the residual sequence of the operating state from the previous sampling period to perform state prediction when it is determined that a non-physical step change occurs in the feedback parameter sequence.

4. The controlled operating component fault monitoring and processing logic method based on multi-source operating data fusion according to claim 1, characterized in that, Step S4 is followed by: extracting trend terms that show a monotonically changing trend from the residual sequence of the operating state, converting the trend terms into the update step size of the time constant of the controlled operating component, and performing online parameter calibration on the preset monitoring reference model.

5. The controlled operating component fault monitoring and processing logic method based on multi-source operating data fusion according to claim 1, characterized in that, Step S5 is followed by: superimposing a high-frequency micro-excitation disturbance signal onto the drive-end current command sequence during the steady-state operation cycle, and obtaining the frequency response characteristics of the feedback parameter sequence to the high-frequency micro-excitation disturbance signal; and extracting structural response indices at specific frequencies. Monitor the structural connection status and structural response indicators of the controlled operating components. The calculation method is as follows: ,in, In frequency The measured impedance amplitude, in units of , This represents the phase response offset angle at the corresponding frequency, in rad. This represents the total number of sampling frequency points.

6. The controlled operating component fault monitoring and processing logic method based on multi-source operating data fusion according to claim 1, characterized in that, In step S1, the feedback parameter sequence includes 50Hz data representing the rotation frequency, 10mm data representing the displacement, and 0.5mm data representing the temperature rise rate. / s data; multi-source runtime data is mapped to a multi-dimensional monitoring space consisting of time, logical state, and execution intensity through a normalization operator.

7. The controlled operating component fault monitoring and processing logic method based on multi-source operating data fusion according to claim 1, characterized in that, In step S3, the dynamic deviation characteristic value is obtained by calculating the time-domain overlap integral between the step rise rate and the response start time, and the calculation step size of the real-time phase offset value is dynamically corrected by using the amplitude of the time-domain overlap integral.

8. The controlled operating component fault monitoring and processing logic method based on multi-source operating data fusion according to claim 1, characterized in that, In step S6, the preset monitoring threshold is calibrated based on the health benchmark database of the controlled operating components; When the residual sequence of the operating state shows a monotonically increasing trend within a time period of 500ms to 2000ms and the slope exceeds the preset rate of change threshold, the controlled operating component is determined to be in a state of performance degradation.

9. The controlled operating component fault monitoring and processing logic method based on multi-source operating data fusion according to claim 1, characterized in that, The fault handling instructions include shutdown instructions for driving controlled operating components into protection mode, fault alarm signals for display on the terminal, and maintenance work order data containing fault characteristic quantities.

10. The controlled operating component fault monitoring and processing logic method based on multi-source operating data fusion according to claim 1, characterized in that, The fault monitoring and handling logic is deployed between the field controller and the decision server to execute instruction compensation when the controlled operating component deviates from its operating parameters.

Images