Energy consumption detection method, system, device and medium based on industrial internet of things

Abstract

The application belongs to the technical field of industrial internet of things, and specifically discloses an energy consumption detection method, system, device and medium based on industrial internet of things, wherein the method comprises the following steps: time sequence alignment is performed on power curve, current waveform and voltage stability data collected by a multi-source sensor according to a time stamp marking technology to obtain a time sequence synchronization data set; frequency domain characteristics of the current waveform in the time sequence synchronization data set are analyzed based on fast Fourier transform to determine whether there is an abnormal frequency component caused by electromagnetic interference, and an interference source device is obtained; the transmission order of real-time calibration instructions is adjusted according to the interference source device and through a priority scheduling algorithm to obtain an optimized instruction sequence; and the power distribution parameters of parallel devices are updated according to the optimized instruction sequence and through an incremental data fusion method. The application aims to solve the problem of inaccurate energy efficiency evaluation caused by electromagnetic interference, voltage fluctuation and energy consumption deviation when multiple devices are cooperatively operated in an industrial scene.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of industrial Internet of Things, and in particular to an energy consumption detection method, system, device and medium based on industrial Internet of Things. BACKGROUND

[0002] Energy consumption evaluation in the field of industrial Internet of Things is crucial for promoting intelligent manufacturing and green production. The core lies in optimizing energy utilization efficiency and improving the sustainable development capability of production systems through real-time monitoring and analysis of multi-device operation data. However, existing solutions often face limitations in complex industrial scenarios, especially when multiple devices are running in parallel. It is difficult to effectively deal with data synchronization and interference problems in dynamic environments, leading to distorted evaluation results and affecting the accuracy of production decisions. The main shortcomings of current methods are the timing conflicts and electromagnetic interference problems when multiple devices are running in coordination. In high-load production scenarios, real-time calibration instruction transmission and data synchronization between devices are difficult to coordinate.

[0003] For example, when devices execute calibration instructions to adjust current waveforms, electromagnetic interference is triggered, affecting the power distribution of neighboring devices, leading to voltage fluctuations and energy consumption data deviation. This interference not only reduces the accuracy of device-level energy consumption benchmarking, but also increases the load pressure of communication protocols due to repeated transmission of calibration instructions. In addition, the delay problem of multi-layer data aggregation further exacerbates the challenge. In large-scale parallel device clusters, device-level energy consumption data needs to be quickly integrated into production line-level and factory-level evaluation. However, the contradiction between high computational complexity and real-time requirements makes it difficult for the system to maintain the stability of evaluation results in dynamic calibration, especially under high-frequency production rhythms, where interference amplification effects are particularly pronounced.

[0004] For example, on an automated production line, voltage fluctuations caused by device calibration may lead to inconsistent energy efficiency evaluation of the entire production line, thereby affecting the formulation of factory-level energy optimization strategies.

[0005] Therefore, how to coordinate real-time calibration and data synchronization of multiple devices running in parallel in a complex electromagnetic environment, reduce the impact of interference on energy consumption evaluation accuracy, and optimize the real-time performance of multi-layer data aggregation, has become a key problem in building an industrial Internet of Things energy consumption multi-dimensional evaluation and detection system. SUMMARY

[0006] The present application provides an energy consumption detection method, system, device and medium based on industrial Internet of Things, aiming to solve the problem of inaccurate energy efficiency evaluation caused by electromagnetic interference, voltage fluctuations and energy consumption deviation when multiple devices are running in coordination in industrial scenarios.

[0007] To solve the above technical problems, the technical solution adopted by the present application is:

[0008] The energy consumption detection method based on industrial Internet of Things comprises the following steps: time sequence alignment of power curve, current waveform and voltage stability data collected by multi-source sensors according to timestamp marking technology to obtain time sequence synchronization dataset; analysis of frequency domain characteristics of current waveform in time sequence synchronization dataset based on fast Fourier transform to determine whether there is abnormal frequency component caused by electromagnetic interference to obtain interference source equipment; transmission order of real-time calibration instructions is adjusted according to interference source equipment and through priority scheduling algorithm to obtain optimized instruction sequence; power distribution parameters of parallel equipment are updated according to optimized instruction sequence and through incremental data fusion method to obtain real-time power distribution scheme; voltage fluctuation abnormality is determined based on real-time power distribution scheme and through real-time monitoring of voltage stability data to obtain execution effect of calibration instructions; if voltage fluctuation abnormality exists, then segmented processing of multi-source sensor data is carried out through dynamic time window analysis technology to obtain local energy consumption deviation characteristics; stable energy consumption evaluation results under high-frequency production rhythm are obtained by smoothing energy consumption data through Kalman filtering algorithm according to local energy consumption deviation characteristics; consistent performance evaluation model is obtained by hierarchical fusion of device-level and factory-level data through distributed data aggregation technology according to stable energy consumption evaluation results.

[0009] In one aspect of the present application, the step of obtaining time sequence synchronization dataset by time sequence alignment of power curve, current waveform and voltage stability data collected by multi-source sensors according to timestamp marking technology comprises:

[0010] Power curve, current waveform and voltage stability data are obtained from multi-source sensors according to timestamp marking technology to generate initial time sequence dataset;

[0011] Time stamps in initial time sequence dataset are aligned based on linear interpolation algorithm to obtain time-synchronized intermediate dataset; if time stamp deviation of intermediate dataset exceeds a preset threshold, then power curve, current waveform and voltage stability data are adjusted through sliding window method to generate deviation-corrected dataset;

[0012] Time sequence characteristics of power curve, current waveform and voltage stability are extracted from deviation-corrected dataset to obtain feature-enhanced dataset;

[0013] Feature-enhanced dataset is processed based on Kalman filtering algorithm to eliminate noise interference and obtain smoothed dataset;

[0014] High-precision synchronization dataset is obtained by optimizing time sequence differences of multi-source data through dynamic time warping algorithm according to smoothed dataset;

[0015] Final time sequence dataset is generated according to high-precision synchronization dataset to determine whether synchronization precision meets a preset threshold requirement.

[0016] In an aspect of the present application, based on the fast Fourier transform analysis of the time sequence synchronization data set of the current waveform frequency domain characteristics, it is judged whether there is an abnormal frequency component caused by electromagnetic interference, and the steps of obtaining the interference source equipment include:

[0017] According to the time sequence data acquisition system, the current waveform original signal is obtained, the baseline drift and high frequency noise are removed through the pretreatment module, and the standardized current waveform data is obtained;

[0018] Based on the fast Fourier transform algorithm, the frequency domain conversion of the standardized current waveform data is carried out, the frequency spectrum amplitude distribution, the phase information and the energy density of each frequency component are obtained;

[0019] According to the frequency spectrum amplitude distribution, the frequency reference template is established, if the detection frequency component exceeds the preset threshold range, it is marked as an abnormal frequency point, and the abnormal frequency component set is obtained;

[0020] The harmonic characteristics of each frequency point in the abnormal frequency component set are analyzed, the interference type attribute is judged, and the electromagnetic interference mode classification result is determined;

[0021] According to the device frequency fingerprint library matching electromagnetic interference mode classification result, if the interference frequency is consistent with the device characteristic frequency, the potential interference source equipment type is determined;

[0022] Synchronization acquisition timestamp information and potential interference source equipment running state data are obtained, the interference occurrence time is verified through time correlation analysis, and the final interference source equipment positioning result is obtained;

[0023] According to the interference source equipment positioning result, the frequency domain interference characteristic report is generated, and the abnormal frequency parameter and the corresponding device identification information are recorded.

[0024] In an aspect of the present application, according to the interference source equipment, and through the priority scheduling algorithm, the transmission order of real-time calibration instruction is adjusted, and the steps of obtaining the optimized instruction sequence include:

[0025] According to the signal strength and frequency range parameters of the interference source equipment in the industrial Internet of things system, and through the signal analysis module, the interference characteristics of each device type are analyzed, and the device interference influence evaluation matrix is obtained;

[0026] According to the interference intensity value in the device interference influence evaluation matrix, and through the weighted scoring mechanism, the priority weight of each calibration instruction is calculated, if the interference intensity exceeds the preset threshold, the priority weight of the corresponding instruction is improved, and the initial priority order table is obtained;

[0027] The initial priority sorting table is processed based on a priority scheduling algorithm, and combined with real-time requirements and response time constraints, if the instruction response time is less than the system preset time window, the current sorting position is maintained to obtain the final instruction scheduling sequence;

[0028] According to the current system available resource capacity, if the resource occupancy rate exceeds the system carrying upper limit, the instruction execution time interval is adjusted to obtain the resource optimization configuration scheme;

[0029] Real-time state information and bandwidth utilization data of the transmission channel are obtained, and the calibration instructions are distributed to different transmission channels through a load balancing algorithm, if the load of a certain channel exceeds the capacity threshold, the instruction transmission path is redistributed, and the channel distribution strategy is determined;

[0030] According to the channel distribution strategy and the instruction scheduling sequence, a timing controller is used to generate accurate instruction sending time stamps, combined with the execution time estimation value of each instruction, the time conflict between instructions is judged, and a time scheduling matrix is obtained;

[0031] According to the timing arrangement in the time scheduling matrix, the optimized calibration instructions are data encapsulated according to the transmission order through the instruction encapsulation module, if the instruction data packet size exceeds the transmission unit limit, the fragmentation processing is carried out, and the final optimized instruction sequence is generated.

[0032] In one aspect of the present application, according to the optimized instruction sequence, and through the incremental data fusion method, the power distribution parameters of the parallel equipment are updated, and the steps of obtaining the real-time power distribution scheme include:

[0033] The real-time running data of the parallel equipment is obtained, the collected data is denoised and standardized by the data preprocessing module, and the standardized running data set is obtained;

[0034] According to the standardized running data set, the incremental data fusion method is used, combined with the device running efficiency and real-time requirements, the power distribution parameters are updated, and the preliminary power distribution scheme is obtained;

[0035] According to the system resource occupancy of the preliminary power distribution scheme, if the resource occupancy rate exceeds the preset threshold, the power distribution parameters are adjusted to obtain the resource optimized power distribution scheme;

[0036] Based on the support vector machine algorithm, the power distribution balance of the resource optimized power distribution scheme is classified and evaluated to determine whether the power distribution scheme meets the balance requirement; if the power distribution scheme meets the balance requirement, the time sequence data of the power distribution is generated, and the power distribution execution plan is obtained;

[0037] According to the power distribution execution plan, the data encapsulation module is used to encapsulate the distribution parameters into control instructions to generate the final real-time power distribution instruction sequence;

[0038] According to the real-time power distribution instruction sequence, the transmission of the real-time monitoring instruction sequence is monitored, and if it is detected that the transmission delay exceeds a preset threshold, the transmission path is re-distributed, and an optimized transmission strategy is determined.

[0039] In an aspect of the present application, based on the real-time power distribution scheme, and through real-time monitoring of voltage stability data, it is judged whether there is an abnormal voltage fluctuation, and the execution effect of the calibration instruction is obtained. The steps include:

[0040] According to the voltage stability data, the voltage fluctuation characteristics are obtained by time series analysis method, and if the voltage fluctuation characteristics exceed the preset threshold, it is judged whether there is an abnormality by abnormality detection algorithm, and the abnormal state is determined;

[0041] According to the abnormal state, a calibration model established in advance is used to generate a calibration instruction, and the calibration instruction is transmitted to the power distribution unit to obtain an execution feedback;

[0042] According to the execution feedback, the power distribution state is extracted in the execution feedback, the execution effect of the calibration instruction is judged by data comparison method, and if the execution effect does not reach the preset standard, the calibration instruction is adjusted according to the power distribution state to obtain an optimized instruction;

[0043] According to the optimized instruction, the power distribution unit is calibrated again to obtain the final execution effect.

[0044] In an aspect of the present application, according to the local energy consumption deviation characteristics, the energy consumption data is smoothed by Kalman filtering algorithm to obtain stable energy consumption evaluation results under high-frequency production rhythm. The steps include:

[0045] The original energy consumption data collected in the production process is obtained, the local energy consumption deviation characteristics are extracted, and the initial energy consumption deviation data set is obtained;

[0046] The initial energy consumption deviation data set is smoothed based on the Kalman filtering algorithm to obtain smoothed energy consumption data;

[0047] According to the smoothed energy consumption data, the energy consumption change trend under high-frequency production rhythm is calculated to obtain an energy consumption trend sequence; if the fluctuation amplitude of the energy consumption trend sequence exceeds the preset threshold, the abnormal points are marked to obtain the marked energy consumption trend sequence;

[0048] Based on the time window analysis of the marked energy consumption trend sequence, the stable energy consumption data is extracted to obtain the stable energy consumption evaluation result;

[0049] According to the stable energy consumption evaluation result, the energy consumption evaluation precision is calculated in combination with the production rhythm frequency to obtain the precision evaluation value;

[0050] The production process monitoring system is updated in real time based on the accuracy evaluation values ​​to generate optimized monitoring parameters.

[0051] In another aspect, the present invention also relates to an energy consumption detection system based on the Industrial Internet of Things, comprising:

[0052] The timing alignment module is used to perform timing alignment on the power curves, current waveforms and voltage stability data collected by multi-source sensors according to the timestamp marking technology, and obtain a timing synchronization dataset;

[0053] The frequency domain analysis module is used to analyze the frequency domain characteristics of the current waveform in the time-series synchronization dataset based on the Fast Fourier Transform, determine whether there are abnormal frequency components caused by electromagnetic interference, and obtain the interference source device.

[0054] The scheduling optimization module is used to adjust the transmission order of real-time calibration commands based on the interference source device and through a priority scheduling algorithm to obtain an optimized command sequence.

[0055] The data fusion module is used to update the power allocation parameters of parallel devices according to the optimization instruction sequence using an incremental data fusion method, so as to obtain a real-time power allocation scheme.

[0056] The voltage monitoring module is used to determine whether there are abnormal voltage fluctuations based on the real-time power distribution scheme and by monitoring the voltage stability data in real time, so as to obtain the execution effect of the calibration command.

[0057] The segmented processing module is used to determine whether the voltage fluctuation is abnormal. If an abnormality is found, the multi-source sensor data is segmented using dynamic time window analysis technology to obtain local energy consumption deviation characteristics.

[0058] The filtering module is used to smooth the energy consumption data based on the local energy consumption deviation characteristics using the Kalman filtering algorithm, so as to obtain the stable energy consumption evaluation results under high-frequency production cycle time.

[0059] The hierarchical aggregation module is used to perform hierarchical fusion of equipment-level and factory-level data based on the stable energy consumption evaluation results, using distributed data aggregation technology to obtain a consistent energy efficiency evaluation model.

[0060] In another aspect, the present invention also relates to an electronic device comprising:

[0061] A memory on which computer programs are stored;

[0062] A processor is used to execute computer programs in memory to implement the aforementioned energy consumption detection method based on the Industrial Internet of Things.

[0063] In another aspect, the present invention also relates to a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements an energy consumption detection method based on the Industrial Internet of Things.

[0064] Compared with the prior art, the present invention has the following beneficial effects:

[0065] This invention achieves time-series synchronization of power curves, current waveforms, and voltage stability data through timestamp technology. It employs Fast Fourier Transform to extract frequency domain features of the current waveform to identify electromagnetic interference sources. Then, a priority scheduling algorithm optimizes the calibration command sequence, and an incremental data fusion method is used to update power allocation parameters in real time, generating an efficient power allocation scheme. Simultaneously, this invention utilizes real-time monitoring and dynamic time window analysis of voltage stability data, combined with a Kalman filter algorithm to smooth energy consumption data, accurately capturing local energy consumption deviations. Finally, a consistent energy efficiency evaluation model is constructed through distributed data aggregation technology. This invention significantly improves the accuracy and stability of energy efficiency assessment for industrial equipment operation, optimizes resource allocation, and reduces the impact of energy consumption fluctuations on production cycle time. Attached Figure Description

[0066] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained from these drawings without creative effort.

[0067] Figure 1 This is a schematic diagram illustrating the steps of the energy consumption detection method based on the Industrial Internet of Things of the present invention;

[0068] Figure 2 This is a schematic diagram of the energy consumption detection method based on the Industrial Internet of Things of the present invention;

[0069] Figure 3 This is a schematic diagram illustrating the composition of the optimized Industrial Internet of Things (IIoT) involved in this invention.

[0070] Figure 4 This is a block diagram illustrating an electronic device according to an embodiment of the present invention.

[0071] In the diagram, 700 is an electronic device, 701 is a processor, 702 is a memory, 703 is a multimedia component, 704 is an I / O interface, and 705 is a communication component. Detailed Implementation

[0072] The present invention will be further described below with reference to embodiments. These embodiments are merely some, not all, of the embodiments described. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the protection scope of the present invention.

[0073] Example 1

[0074] Please see Figures 1-4 As shown, this embodiment discloses an energy consumption detection method based on the Industrial Internet of Things (IIoT), comprising: performing time-series alignment of power curves, current waveforms, and voltage stability data collected by multi-source sensors using timestamp marking technology to obtain a time-series synchronization dataset; analyzing the frequency domain characteristics of the current waveforms in the time-series synchronization dataset using Fast Fourier Transform to determine whether there are abnormal frequency components caused by electromagnetic interference, thereby identifying the interference source device; adjusting the transmission order of real-time calibration commands according to the interference source device and using a priority scheduling algorithm to obtain an optimized command sequence; and updating the power allocation of parallel devices according to the optimized command sequence and using an incremental data fusion method. The system obtains real-time power allocation schemes based on parameters. Based on these schemes, and through real-time monitoring of voltage stability data, it determines whether voltage fluctuations are abnormal and assesses the effectiveness of calibration commands. If voltage fluctuations are abnormal, it uses dynamic time window analysis to segment the multi-source sensor data, obtaining local energy consumption deviation characteristics. Based on these characteristics, it uses a Kalman filter algorithm to smooth the energy consumption data, obtaining stable energy consumption evaluation results under high-frequency production cycles. Based on these stable energy consumption evaluation results, it employs distributed data aggregation technology to perform layered fusion of equipment-level and factory-level data, obtaining a consistent energy efficiency evaluation model.

[0075] Among them, the algorithm for the Fast Fourier Transform of discrete-time series is used for:

[0076]

[0077] in, This represents a discrete time-domain current waveform signal sequence, where... For sequence indexing ;

[0078] Indicates the length of the time-series signal sequence;

[0079] This represents the frequency domain signal sequence obtained after Fast Fourier Transform, where Frequency domain index ;

[0080] Represents the imaginary unit ;

[0081] e -j2πkn / N It is a complex exponential function, representing a complex rotation factor, used to convert time-domain signals to the frequency domain;

[0082] And e is the natural constant;

[0083] This embodiment converts the time-domain signal of the current waveform into a frequency-domain signal using the above formula, and can extract features such as spectral amplitude distribution and phase information, providing a quantitative basis for identifying abnormal frequency components of electromagnetic interference.

[0084] In practical applications, this invention achieves time-series synchronization of power curves, current waveforms, and voltage stability data through timestamp technology. It employs Fast Fourier Transform to extract frequency domain features of the current waveform to identify electromagnetic interference sources. Furthermore, it optimizes the calibration command sequence using a priority scheduling algorithm and updates power allocation parameters in real time using incremental data fusion to generate an efficient power allocation scheme. Simultaneously, this invention utilizes real-time monitoring and dynamic time window analysis of voltage stability data, combined with Kalman filtering to smooth energy consumption data, accurately capturing local energy consumption deviations. Finally, it constructs a consistent energy efficiency evaluation model through distributed data aggregation technology. This invention significantly improves the accuracy and stability of energy efficiency assessment for industrial equipment operation, optimizes resource allocation, and reduces the impact of energy consumption fluctuations on production cycle time.

[0085] Example 2

[0086] Please see Figures 1-4 As shown, this embodiment is a further optimization based on Embodiment 1. In this embodiment, the step of performing time-series alignment on the power curves, current waveforms, and voltage stability data collected by multiple source sensors according to timestamp marking technology to obtain a time-series synchronization dataset includes:

[0087] Power curves, current waveforms, and voltage stability data are acquired from multi-source sensors using timestamp technology to generate an initial time series dataset.

[0088] The timestamps in the initial time series dataset are aligned using a linear interpolation algorithm to obtain a time-synchronized intermediate dataset. If the timestamp deviation of the intermediate dataset exceeds a preset threshold, the power curve, current waveform, and voltage stability data are adjusted using a sliding window method to generate a deviation-corrected dataset.

[0089] The interpolation results in the linear interpolation algorithm are used for:

[0090]

[0091] in, , This represents two known timestamps;

[0092] This represents the target timestamp for which interpolation calculations are required, where t is between t0 and t1;

[0093] This represents the interpolation result at the target timestamp t, i.e., the data value obtained through linear interpolation;

[0094] , Indicates the corresponding timestamp;

[0095] , The collected data values;

[0096] The above formula assumes two known data points. as well as By calculating the unknown data values ​​between two known time points through linear relationships, the initial alignment of timestamps can be quickly achieved, enabling time-series synchronization preprocessing of multi-source sensor data.

[0097] Based on the bias-corrected dataset, time-series features of power curves, current waveforms, and voltage stability are extracted to obtain a feature-enhanced dataset.

[0098] The feature enhancement dataset is processed using the Kalman filter algorithm to eliminate noise interference and obtain a smoothed dataset.

[0099] The Kalman filter algorithm used for data smoothing consists of two steps, as follows:

[0100] 1. Prediction:

[0101] 2. Update:

[0102] in, Indicates in Prior estimates at time (estimates obtained during the prediction phase);

[0103] This represents the state transition matrix, used to describe the dynamic changes in the system state;

[0104] Indicates in The posterior estimate at time (the estimate obtained during the update phase);

[0105] This represents the control input matrix, used to associate control inputs with state transitions;

[0106] Indicates again Time-based control input;

[0107] Indicates in The prior estimate error covariance matrix at time t;

[0108] Indicates in The posterior estimation error covariance matrix at time t;

[0109] Representation matrix transpose;

[0110] This represents the process noise covariance matrix, used to describe the uncertainty of the system model;

[0111] Indicates again The Kalman gain at time step is used to weigh the reliability of prior estimates and observations;

[0112] This represents the observation matrix, used to associate the system state with the observed values;

[0113] Representation matrix transpose;

[0114] This represents the noise covariance matrix, used to describe the uncertainty of the observation process;

[0115] Representation matrix The inverse matrix;

[0116] Indicates in The posterior estimate at time (the final estimate obtained during the update phase);

[0117] Indicates in The observed value at time;

[0118] Indicates in The posterior estimation error covariance matrix at time t;

[0119] Represents the identity matrix.

[0120] The above formula can effectively eliminate noise interference in energy consumption data and achieve smooth processing of energy consumption data under high-frequency production cycle time.

[0121] Based on the smoothed dataset, a dynamic time warping algorithm is used to optimize the temporal differences of multi-source data to obtain a high-precision synchronized dataset.

[0122] In the dynamic time warping algorithm, firstly, two time series are set up. as well as Its dynamic time warping is used as an example for:

[0123]

[0124] in, This represents the first time series, with a length of [length missing]. ,in For the first in the sequence One element;

[0125] This represents the second time series, with a length of [length missing]. ,in For the first in the sequence One element;

[0126] Representing time series and The dynamic time-normalized distance between two sequences is used to measure their similarity.

[0127] Representing time series The Middle element With time series The Middle element The Euclidean distance between them;

[0128] Represents the path weight, used to indicate the weight of a path in a dynamically time-warped path. The weights traversed by a location must satisfy certain constraints (such as monotonicity, continuity, etc.).

[0129] The above formula can optimize the temporal differences of multi-source sensor data and improve synchronization accuracy by finding the optimal matching path to align time series of different lengths.

[0130] Based on the high-precision synchronization dataset, the final time-series dataset is generated, and it is determined whether the synchronization accuracy meets the preset threshold requirements.

[0131] In practical application, this embodiment first acquires relevant data from multiple sensors using timestamp marking technology to generate an initial time-series dataset. Then, a linear interpolation algorithm is used to align the timestamps to obtain an intermediate dataset. If the timestamp deviation exceeds a threshold, a sliding window method is used to adjust and generate a deviation-corrected dataset. Next, time-series features are extracted to obtain a feature-enhanced dataset. Noise is eliminated using Kalman filtering to obtain a smoothed dataset. Then, a dynamic time warping algorithm is used to optimize time-series differences to obtain a high-precision synchronized dataset. Finally, the final time-series dataset is generated, and the synchronization accuracy is assessed. This embodiment solves the problem of time-series asynchrony in data acquired from multiple sensors. Through a series of data processing and optimization steps, it improves the synchronization accuracy of the data, providing an accurate and consistent time-series data foundation for subsequent energy consumption detection and analysis, ensuring the reliability of subsequent detection results.

[0132] In some embodiments, the step of analyzing the frequency domain characteristics of the current waveform in the time-series synchronization dataset based on Fast Fourier Transform to determine whether there are abnormal frequency components caused by electromagnetic interference, and obtaining the interference source device, includes:

[0133] Based on the time-series data acquisition system, the raw current waveform signal is acquired, and the baseline drift and high-frequency noise are removed by the preprocessing module to obtain standardized current waveform data.

[0134] The standardized current waveform data is frequency domain transformed based on the Fast Fourier Transform algorithm to obtain the spectral amplitude distribution, phase information, and energy density of each frequency component.

[0135] A frequency reference template is established based on the spectral amplitude distribution. If the detected frequency component exceeds the preset threshold range, it is marked as an abnormal frequency point, and a set of abnormal frequency components is obtained.

[0136] Analyze the harmonic characteristics of each frequency point in the abnormal frequency component set, determine the interference type attribute, and determine the electromagnetic interference mode classification result;

[0137] Based on the electromagnetic interference pattern classification results matched with the device frequency fingerprint database, if the interference frequency matches the device characteristic frequency, the type of potential interference source device is determined.

[0138] Acquire synchronous collection timestamp information and operating status data of potential interference source devices, verify the time of interference occurrence through time correlation analysis, and obtain the final interference source device location result;

[0139] A frequency domain interference characteristic report is generated based on the location results of the interference source device, recording abnormal frequency parameters and corresponding device identification information.

[0140] In practical use, this embodiment acquires the raw current waveform signal through a time-series data acquisition system. A preprocessing module removes baseline drift and high-frequency noise to obtain standardized current waveform data. A Fast Fourier Transform (FFT) algorithm is used for frequency domain conversion. A frequency reference template is established based on the spectral amplitude distribution. Abnormal frequency points exceeding thresholds are marked. The harmonic characteristics of the abnormal frequency components are analyzed to determine the electromagnetic interference mode classification results. A device frequency fingerprint database is used to match the type of potential interference source device. Finally, time correlation analysis verifies the time of interference occurrence to obtain the final interference source device location result, and a frequency domain interference feature report is generated. This embodiment solves the problem of difficult identification of electromagnetic interference sources in industrial scenarios. By analyzing the frequency domain characteristics of the current waveform and locating the interference source, the interference source device causing the anomaly can be accurately identified, providing a basis for subsequent targeted measures to eliminate interference and reducing the impact of electromagnetic interference on the accuracy of energy consumption detection.

[0141] In some embodiments, the step of obtaining an optimized instruction sequence by adjusting the transmission order of real-time calibration instructions based on the interference source device and using a priority scheduling algorithm includes:

[0142] Based on the signal strength and frequency range parameters of interference source devices in the industrial Internet of Things system, and by analyzing the interference characteristics of each device type through the signal analysis module, an equipment interference impact assessment matrix is ​​obtained.

[0143] Based on the interference intensity values ​​in the equipment interference impact assessment matrix, and through a weighted scoring mechanism, the priority weight of each calibration command is calculated. If the interference intensity exceeds the preset threshold, the priority weight of the corresponding command is increased to obtain the initial priority ranking table.

[0144] The initial priority sorting table is processed based on the priority scheduling algorithm. Combining real-time requirements and response time constraints, if the instruction response time is less than the system's preset time window, the current sorting position is maintained to obtain the final instruction scheduling order.

[0145] Based on the current available system resources, if the resource utilization rate exceeds the system's capacity limit, the instruction execution time interval is adjusted to obtain a resource optimization configuration scheme.

[0146] The system acquires real-time status information and bandwidth utilization data of the transmission channels, distributes calibration commands to different transmission channels through a load balancing algorithm, and reallocates command transmission paths and determines channel allocation strategies if the load of a channel exceeds the capacity threshold.

[0147] The load balancing algorithm is used to calibrate the allocation of instruction transmission channels; in the load balancing algorithm, let... One calibration instruction was assigned to The transmission channel, the first j The load factor of each channel is used for:

[0148]

[0149] in, Indicates the total number of calibration instructions that need to be assigned;

[0150] Indicates the number of transmission channels;

[0151] Index representing the transmission channel ;

[0152] Indicates the first The load factor of a transmission channel is the ratio of the actual load to the capacity of that channel.

[0153] Indicates allocation to the first A set of calibration instructions for each transmission channel;

[0154] Indicates the first The workload of each calibration command, such as data size and processing time;

[0155] Indicates the first The capacity of each transmission channel, such as: maximum data processing capacity, maximum bandwidth;

[0156] Indicates allocation to the first The total load of all calibration commands for each transmission channel;

[0157] The above formula, by minimizing Achieve load balancing, ensure that the load on each channel does not exceed the capacity threshold, and optimize command transmission paths;

[0158] in, This represents the maximum load rate among all transmission channels, and the goal of the load balancing algorithm is to minimize this value.

[0159] Based on the channel allocation strategy and instruction scheduling order, a timing controller is used to generate accurate instruction sending timestamps. Combined with the estimated execution time of each instruction, the time conflict between instructions is determined, and a time scheduling matrix is ​​obtained.

[0160] According to the timing arrangement in the time scheduling matrix, the optimized calibration instructions are encapsulated in the transmission order by the instruction encapsulation module. If the instruction data packet size exceeds the transmission unit limit, it is fragmented to generate the final optimized instruction sequence.

[0161] In practical use, this embodiment analyzes interference characteristics based on parameters such as signal strength and frequency range of the interference source device to obtain a device interference impact assessment matrix. Based on the interference strength values ​​in this matrix, the priority weights of calibration instructions are calculated, and an initial priority ranking table is generated. Combining real-time requirements and response time constraints, the final instruction scheduling order is obtained. The instruction execution interval and transmission path are adjusted according to system resource capacity and transmission channel status. After generating a time scheduling matrix, calibration instructions are sequentially encapsulated. If the data packet is too large, it is fragmented to generate an optimized instruction sequence. This embodiment solves the problem of unreasonable real-time calibration instruction transmission order. Through a priority scheduling algorithm and optimized configuration of resources and transmission channels, calibration instructions can be transmitted and executed more efficiently and orderly, improving the real-time performance and reliability of instruction execution and reducing the load on the communication protocol.

[0162] In some embodiments, the step of obtaining a real-time power allocation scheme by updating the power allocation parameters of parallel devices according to an optimized instruction sequence and through an incremental data fusion method includes:

[0163] The system acquires real-time operating data of parallel devices, and performs noise reduction and standardization on the collected data through a data preprocessing module to obtain a standardized operating dataset.

[0164] Based on the standardized operational dataset, an incremental data fusion method is used to update the power allocation parameters and obtain a preliminary power allocation scheme by combining equipment operating efficiency and real-time requirements.

[0165] Based on the system resource usage of the initial power allocation scheme, if the resource utilization rate exceeds the preset threshold, the power allocation parameters are adjusted to obtain a resource-optimized power allocation scheme.

[0166] The power allocation balance of the resource optimization power allocation scheme is classified and evaluated based on the support vector machine algorithm to determine whether the power allocation scheme meets the balance requirement. If the power allocation scheme meets the balance requirement, time series data of power allocation is generated to obtain the power allocation execution plan.

[0167] In the Support Vector Machine (SVM) algorithm, for binary classification problems, the optimal classification hyperplane is used for:

[0168]

[0169] in, This represents the classification function of a support vector machine, used to classify input samples. Perform classification prediction;

[0170] This represents a sign function. When the value inside the parentheses is positive, the function value is 1; when the value inside the parentheses is negative, the function value is -1; and when the value inside the parentheses is 0, the function value is 0.

[0171] This represents the weight vector, used to determine the orientation of the classification hyperplane;

[0172] Represents the weight vector transpose;

[0173] This represents the input sample feature vector;

[0174] This represents the bias term, used to determine the position of the classification hyperplane;

[0175] in Used for:

[0176]

[0177] in, This represents the penalty coefficient, used to balance the training error and generalization ability of the model. The larger the value, the heavier the penalty for misclassified samples;

[0178] Indicates the number of training samples;

[0179] This represents a slack variable, used to allow for the possibility of misclassification of some samples. ≥0;

[0180] Represents the weight vector of The square of the norm is used to measure the complexity of the model;

[0181] The above formula can be used to classify and evaluate the balance of power distribution schemes and determine whether they meet the preset requirements.

[0182] According to the power allocation execution plan, the data encapsulation module encapsulates the allocation parameters into control commands to generate the final real-time power allocation command sequence.

[0183] Based on the real-time power allocation command sequence, the transmission status of the command sequence is monitored in real time. If the transmission delay is detected to exceed the preset threshold, the transmission path is reallocated and an optimized transmission strategy is determined.

[0184] In some embodiments, real-time operating data of parallel devices is acquired, preprocessed to obtain a standardized operating dataset, and an incremental data fusion method is used to update power allocation parameters to obtain a preliminary power allocation scheme. If the resource occupancy rate exceeds a preset threshold, the parameters are adjusted to obtain a resource-optimized power allocation scheme. The balance of the scheme is evaluated using a support vector machine algorithm. If the requirements are met, a power allocation execution plan is generated, encapsulated into control commands to generate a real-time power allocation command sequence, and the transmission status is monitored in real time. If the delay exceeds a threshold, the transmission path is reallocated. The above embodiments solve the problem of unreasonable power allocation of parallel devices. Through incremental data fusion and a series of optimization and adjustment steps, real-time updating and optimization of power allocation parameters are achieved, improving the balance of power allocation and resource utilization of parallel devices, and ensuring the high efficiency and stability of device operation.

[0185] In some embodiments, the step of determining whether there are abnormal voltage fluctuations based on a real-time power allocation scheme and through real-time monitoring of voltage stability data, and obtaining the execution effect of the calibration command, includes:

[0186] Based on voltage stability data, voltage fluctuation characteristics are obtained through time series analysis. If the voltage fluctuation characteristics exceed a preset threshold, an anomaly detection algorithm is used to determine whether an anomaly exists and to identify the abnormal state.

[0187] Based on the abnormal state, a pre-established calibration model is used to generate calibration instructions, which are then transmitted to the power distribution unit to obtain execution feedback.

[0188] Based on the execution feedback, the power allocation status is extracted from the execution feedback, and the execution effect of the calibration command is judged by the data comparison method. If the execution effect does not meet the preset standard, the calibration command is adjusted according to the power allocation status to obtain the optimized command.

[0189] The power distribution unit is calibrated a second time according to the optimization instructions to obtain the final execution effect.

[0190] In practical use, this embodiment obtains voltage fluctuation characteristics through time series analysis based on voltage stability data. If the voltage exceeds a threshold, it is judged as an abnormal state. A pre-established calibration model is used to generate calibration instructions and transmit them to the power distribution unit. The execution effect of the instructions is judged based on the execution feedback. If the preset standard is not met, the calibration instructions are adjusted for secondary calibration to obtain the final execution effect. The above embodiment solves the problems of abnormal voltage fluctuations and poor calibration instruction execution effect. By real-time monitoring of voltage stability data and dynamic adjustment of calibration instructions, abnormal voltage fluctuations can be detected and corrected in a timely manner, improving the execution effect of calibration instructions, ensuring voltage stability, and thus reducing the impact of voltage fluctuations on equipment operation and energy consumption.

[0191] In some embodiments, if voltage fluctuations are abnormal, the step of segmenting multi-source sensor data using dynamic time window analysis technology to obtain local energy consumption deviation characteristics includes:

[0192] If abnormal voltage fluctuations are detected, the multi-source sensor data is filtered through a preset threshold to obtain abnormal data segments.

[0193] By using dynamic time window analysis, abnormal data segments are segmented and local energy consumption deviation characteristics are extracted.

[0194] Principal component analysis algorithm is used to reduce the dimensionality of local energy consumption deviation features and obtain the main feature vectors;

[0195] If the magnitude of the main feature vector exceeds the preset threshold, the feature vector is classified using the K-means clustering algorithm to determine the anomaly category.

[0196] Based on the anomaly category, time series analysis is used to predict the trend of energy consumption changes by classifying the feature vectors.

[0197] The energy consumption change trend is matched with a preset rule base to determine the type of anomaly.

[0198] If the anomaly source type is equipment failure, then deep analysis of multi-source sensor data is performed through association rule mining to obtain the faulty equipment identifier.

[0199] In some embodiments, the step of smoothing energy consumption data using a Kalman filter algorithm based on local energy consumption deviation characteristics to obtain stable energy consumption evaluation results under high-frequency production cycle time includes:

[0200] Obtain raw energy consumption data collected during the production process, extract local energy consumption deviation features, and obtain an initial energy consumption deviation dataset;

[0201] The initial energy consumption deviation dataset is smoothed using the Kalman filter algorithm to obtain smoothed energy consumption data.

[0202] Based on smoothed energy consumption data, the energy consumption trend under high-frequency production cycle is calculated to obtain an energy consumption trend sequence. If the fluctuation range of the energy consumption trend sequence exceeds a preset threshold, the outliers are marked to obtain a marked energy consumption trend sequence.

[0203] Based on time window analysis of the labeled energy consumption trend series, stable energy consumption data is extracted to obtain stable energy consumption evaluation results;

[0204] Based on the stable energy consumption assessment results and combined with the production cycle frequency, the energy consumption assessment accuracy is calculated to obtain the accuracy assessment value;

[0205] The production process monitoring system is updated in real time based on the accuracy evaluation values ​​to generate optimized monitoring parameters.

[0206] In practical application, this embodiment obtains an initial energy consumption deviation dataset by acquiring raw energy consumption data and extracting local energy consumption deviation features. It then uses a Kalman filter algorithm for smoothing to obtain smoothed energy consumption data. The energy consumption trend under high-frequency production cycles is calculated, and outliers with fluctuations exceeding thresholds are marked. Time window analysis is performed on the marked energy consumption trend sequence to extract stable energy consumption data. The accuracy of energy consumption assessment is calculated in conjunction with the production cycle frequency, and the parameters of the production process monitoring system are updated accordingly. This embodiment solves the problem of large fluctuations in energy consumption data and unstable assessment results under high-frequency production cycles. Through Kalman filtering smoothing and time window analysis, stable energy consumption assessment results are obtained, improving the accuracy and stability of energy consumption assessment and providing an accurate basis for factories to formulate energy optimization strategies.

[0207] In some different embodiments, this embodiment also relates to an energy consumption detection system based on the Industrial Internet of Things, comprising:

[0208] The timing alignment module is used to perform timing alignment on the power curves, current waveforms and voltage stability data collected by multi-source sensors according to the timestamp marking technology, and obtain a timing synchronization dataset;

[0209] The frequency domain analysis module is used to analyze the frequency domain characteristics of the current waveform in the time-series synchronization dataset based on the Fast Fourier Transform, determine whether there are abnormal frequency components caused by electromagnetic interference, and obtain the interference source device.

[0210] The scheduling optimization module is used to adjust the transmission order of real-time calibration commands based on the interference source device and through a priority scheduling algorithm to obtain an optimized command sequence.

[0211] The data fusion module is used to update the power allocation parameters of parallel devices according to the optimization instruction sequence using an incremental data fusion method, so as to obtain a real-time power allocation scheme.

[0212] The voltage monitoring module is used to determine whether there are abnormal voltage fluctuations based on the real-time power distribution scheme and by monitoring the voltage stability data in real time, so as to obtain the execution effect of the calibration command.

[0213] The segmented processing module is used to determine whether the voltage fluctuation is abnormal. If an abnormality is found, the multi-source sensor data is segmented using dynamic time window analysis technology to obtain local energy consumption deviation characteristics.

[0214] The filtering module is used to smooth the energy consumption data based on the local energy consumption deviation characteristics using the Kalman filtering algorithm, so as to obtain the stable energy consumption evaluation results under high-frequency production cycle time.

[0215] The hierarchical aggregation module is used to perform hierarchical fusion of equipment-level and factory-level data based on the stable energy consumption evaluation results, using distributed data aggregation technology to obtain a consistent energy efficiency evaluation model.

[0216] In practical applications, this energy consumption monitoring system based on the Industrial Internet of Things (IIoT) achieves energy consumption monitoring through the collaborative work of multiple modules. First, the timing alignment module uses timestamp technology to align the power curves, current waveforms, and voltage stability data collected by multiple sensors, obtaining a timing synchronization dataset. The frequency domain analysis module analyzes the frequency domain characteristics of the current waveforms in the timing synchronization dataset using Fast Fourier Transform (FFT) to identify and locate interfering devices. The scheduling optimization module adjusts the transmission order of real-time calibration commands based on the interfering devices using a priority scheduling algorithm, obtaining an optimized command sequence. The data fusion module updates the power allocation parameters of parallel devices using an incremental data fusion method based on the optimized command sequence, obtaining a real-time power allocation scheme. The voltage monitoring module... In the real-time power allocation scheme, voltage fluctuation anomalies are detected through real-time monitoring of voltage stability data, and the calibration command execution effect is obtained. When voltage fluctuation anomalies occur, the segmented processing module uses dynamic time window analysis technology to segment the multi-source sensor data to obtain local energy consumption deviation characteristics. Based on the local energy consumption deviation characteristics, the filtering processing module uses the Kalman filter algorithm to smooth the energy consumption data and obtain stable energy consumption evaluation results under high-frequency production cycle. Based on the stable energy consumption evaluation results, the hierarchical aggregation module uses distributed data aggregation technology to perform hierarchical fusion of equipment-level and factory-level data to obtain a consistent energy efficiency evaluation model.

[0217] The system primarily addresses a range of issues arising from the collaborative operation of multiple devices in industrial settings. These include: asynchronous data acquisition from multiple sensors leading to distorted analysis and detection results; difficulty in identifying electromagnetic interference sources, affecting the accuracy of energy consumption detection; unreasonable transmission order of real-time calibration commands, reducing command execution efficiency; untimely and unbalanced updates of power allocation parameters for parallel devices, impacting equipment operating efficiency; difficulty in real-time monitoring and calibration of abnormal voltage fluctuations; large fluctuations in energy consumption data and unstable evaluation results under high-frequency production cycles; and inconsistencies between equipment-level and factory-level data fusion, affecting the accuracy of energy efficiency assessments.

[0218] The advantages of this setup are that, through the collaborative work of its modules, the system significantly improves the accuracy and stability of energy efficiency assessment for industrial equipment operation. It achieves precise time-series synchronization of multi-source sensor data, providing a reliable data foundation for subsequent testing; accurately identifies electromagnetic interference sources, reducing the impact of interference on testing accuracy; optimizes the transmission and execution of calibration commands, improving real-time performance and reliability; achieves reasonable power allocation for parallel equipment, improving resource utilization; effectively monitors and corrects voltage fluctuations, ensuring stable equipment operation; obtains stable energy consumption assessment results under high-frequency production cycles, providing an accurate basis for energy optimization strategies; and through hierarchical fusion of equipment-level and factory-level data, obtains a consistent energy efficiency assessment model, helping factories formulate scientific and reasonable energy management strategies, reducing the impact of energy consumption fluctuations on production cycles, and promoting intelligent manufacturing and green production.

[0219] As an optional implementation method, this embodiment provides one optional implementation method. Of course, the implementation method in this embodiment is not limited to this one embodiment. In this embodiment:

[0220] For example, in the automated production workshop of a smart manufacturing factory, ten production lines operate in parallel, each equipped with 20 welding robots, and all equipment is connected to an industrial Internet of Things (IoT) system. To achieve precise energy consumption management, the system operates according to the following process:

[0221] The timing alignment module starts working first, collecting data from multi-source sensors distributed across various devices. At 10:00:00, sensor A collects a power value of 200kW for welding robot #1, and sensor B collects current waveform data for the same robot at 10:00:02.

[0222] Because of a 2-second timestamp discrepancy, the system calls a linear interpolation algorithm formula, where the data collected from the multi-source sensors at the target device... Power value at 10s =200 , Power value at =12s =220 , needs to be calculated =Power value at 11s (that is, between 10s and 12s);

[0223] Substitute the formula:

[0224] Therefore, when the target device is Power value at 10s =200 , Power value at =12s =220 The power value at that time was 210. This allows for the initial alignment of timestamps;

[0225] When the timestamp deviation of a certain set of data is detected to reach 60ms (exceeding the 50ms threshold), a sliding window (window size 1s) is started for secondary correction. Finally, the data synchronization accuracy of all devices is controlled within 8ms through the dynamic time warping algorithm, and a time-series synchronization dataset is generated.

[0226] In this process, the time-domain discrete signal x[n]=[1, 3, 5, 7] (sampling frequency 50Hz) of the target device current waveform is preprocessed to remove high-frequency noise above 50Hz, and then converted to the frequency domain using the fast Fourier transform formula;

[0227] The process of substituting into the Fast Fourier Transform formula is as follows: When When =0, ;

[0228] when When =1, ;

[0229] e -j2πkn / N It is a complex exponential function, representing a complex rotation factor, used to convert time-domain signals to the frequency domain;

[0230] When K=0, it is naturally represented as e. 0 ;

[0231] And e is the natural constant;

[0232] Calculations show that when When =1, the following occurs The abnormal frequency component was matched with the characteristic frequency of 150Hz of the welding robot in the equipment frequency fingerprint database. Combined with timestamp analysis, it was confirmed that robot #3 had electromagnetic interference between 10:00:05 and 10:00:10, and an interference report containing abnormal frequency parameters was generated.

[0233] The scheduling optimization module initiates priority scheduling based on the interference report and constructs an equipment interference impact assessment matrix. The interference intensity of robot #3 reaches -45dBm (exceeding the -50dBm threshold), and the priority weight of the corresponding calibration command is increased from 0.3 to 0.8. The system calls the load balancing algorithm to distribute the four calibration commands ( ) allocated to 3 transmission channels (capacity) ), through formula Calculate the channel load rate, if allocated ;

[0234] but ; ; ;

[0235] Therefore, the load rate of all three channels exceeded the 80% capacity threshold, and a time scheduling matrix was finally generated to ensure that instructions were responded to within 100ms.

[0236] After receiving the optimization command sequence, the data fusion module collects real-time operating data of the parallel devices (current 30A, voltage 380V, etc.), and updates the power allocation parameters using incremental fusion after noise reduction. The initial scheme shows that the resource utilization rate of robot #5 reaches 92% (exceeding the 90% threshold). After adjustment, the power allocation balance σ value is evaluated by the support vector machine algorithm and is 0.04 (meeting the standard of <0.05). An execution plan containing power adjustment values ​​every 100ms is generated and encapsulated into a control command sequence for transmission.

[0237] The voltage monitoring module tracks voltage stability data in real time and detects a voltage fluctuation of 7% at 10:02:15 (exceeding the 5% threshold), immediately initiating anomaly detection. An initial command is generated based on the PID calibration model, and execution feedback shows a power deviation of 3.2%. After adjusting parameters and performing a second calibration, the fluctuation is ultimately controlled to 2.1%.

[0238] The segmentation processing module segments the data during periods of voltage anomalies, extracts local energy consumption deviation characteristics through a dynamic time window, and the filtering processing module uses a Kalman filter algorithm for smoothing. The initial estimate is set to 100. Observation value 105 According to the prediction formula and updated formula Perform calculations;

[0239] In the target equipment energy consumption system, the state transition matrix =1, control input =0, observation matrix H=1, process noise covariance =0.1, observation noise covariance R=0.5, initial estimate Initial estimation error covariance Observation at time 1 =150 After substituting into the formula:

[0240]

[0241]

[0242]

[0243]

[0244] Therefore, the final smoothed energy consumption data was 103.4kW. Combined with a production cycle of 60 times / minute, a stable energy consumption sequence was extracted through time window analysis, with an accuracy evaluation value of 98.2%.

[0245] Finally, the hierarchical aggregation module merges equipment-level data (energy consumption of a single robot) with factory-level data (total energy consumption of the production line) to build a consistent energy efficiency evaluation model.

[0246] The system showed that after implementation, the workshop's energy consumption fluctuation decreased from ±15% to ±5%, and the average daily power saving per production line was 86.4 kWh, providing accurate data support for the factory's energy optimization.

[0247] It should also be noted that the entire data protocol conversion system based on the Industrial Internet of Things (IIoT) can be applied to the optimized IIoT, such as... Figure 3 As shown, the optimized Industrial Internet of Things (IIoT) includes a user platform, a service platform, a management platform, a sensor network platform, and an object platform that establish communication in sequence.

[0248] The user platform is used to provide front-end services to users; users obtain the necessary perception service information through the user platform, process the perception service information, and transform it into user perception information; users analyze the user perception information and make corresponding decisions based on their own wishes, and transform the user perception information into user control information through the corresponding information system and send it to the service platform, thereby demonstrating the user's corresponding service needs and wishes.

[0249] The physical entities of the user platform include various user terminals, such as mobile phones, computers, and dedicated terminals, which provide user services through integration with user information system software.

[0250] The service platform is used for API servers or other servers to establish communication between the management platform and the user platform to achieve corresponding functions; the physical entities of the service platform include various servers.

[0251] The management platform is used for at least one of the following: equipment operation status monitoring and management, data monitoring and management, equipment parameter management, and lifecycle management; the management platform is an IoT data classification platform, which may include various management sub-platforms, and different management sub-platforms perform different management functions; the physical entity of the management platform includes various servers.

[0252] The sensor network platform is used for at least one of the following: network management, command management, device status management, data protocol management, data parsing, data classification, data transmission monitoring, and data transmission security management. The sensor network platform provides functions such as data communication, transmission, parsing, identification, and classification, avoiding the direct aggregation of data from various object platforms on the management platform, which would result in data redundancy and low data processing efficiency. The physical entities of the object platforms include various gateways, edge computing devices, etc.

[0253] The object platform is used for specific production control, testing, measurement and other production work; the physical entities of the production objects include various production equipment, sensors and so on.

[0254] Figure 4 This is a block diagram of an electronic device for an energy consumption detection method based on the Industrial Internet of Things (IIoT) according to an exemplary embodiment. Figure 4 As shown, the electronic device 700 may include: a processor 701 and a memory 702. The electronic device 700 may also include one or more of a multimedia component 703, an I / O interface 704 (input / output interface), and a communication component 705.

[0255] The processor 701 controls the overall operation of the electronic device 700 to complete all or part of the steps in the aforementioned energy consumption detection method based on the Industrial Internet of Things. The memory 702 stores various types of data to support the operation of the electronic device 700. This data may include, for example, instructions for any application or method operating on the electronic device 700, and application-related data such as contact data, sent and received messages, images, audio, video, etc. The memory 702 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read-Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. Multimedia component 703 may include a screen and an audio component. The screen may be, for example, a touchscreen, and the audio component is used to output and / or input audio signals. For example, the audio component may include a microphone for receiving external audio signals. The received audio signals may be further stored in memory 702 or transmitted via communication component 705. The audio component also includes at least one speaker for outputting audio signals. I / O interface 704 provides an interface between processor 701 and other interface modules, such as a keyboard, mouse, buttons, etc. These buttons may be virtual or physical buttons. Communication component 705 is used for wired or wireless communication between the electronic device 700 and other devices. Wireless communication, such as Wi-Fi, Bluetooth, Near Field Communication (NFC), 2G, 3G, 4G, NB-IoT, eMTC, or other 5G technologies, or combinations thereof, is not limited here. Therefore, the corresponding communication component 705 may include: a Wi-Fi module, a Bluetooth module, an NFC module, etc.

[0256] In an exemplary embodiment, the electronic device 700 may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the above-described energy consumption detection method based on the Industrial Internet of Things.

[0257] In another exemplary embodiment, a computer-readable storage medium including program instructions is also provided. When executed by a processor, these program instructions implement the steps of the above-described energy consumption detection method based on the Industrial Internet of Things (IIoT). For example, the computer-readable storage medium may be the memory 702 including the program instructions, which may be executed by the processor 701 of the electronic device 700 to complete the above-described energy consumption detection method based on the IIoT.

[0258] In another exemplary embodiment, a computer program product is also provided, the computer program product comprising a computer program executable by a programmable device, the computer program having a code portion for performing the above-described energy consumption detection method based on the Industrial Internet of Things when executed by the programmable device.

[0259] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An energy consumption detection method based on the Industrial Internet of Things, characterized in that, include: The power curves, current waveforms, and voltage stability data collected by multiple source sensors are time-aligned using timestamp marking technology to obtain a time-synchronized dataset. Based on the analysis of the frequency domain characteristics of the current waveform in the time-series synchronization dataset using Fast Fourier Transform, it is determined whether there are abnormal frequency components caused by electromagnetic interference, and the interference source device is obtained. Based on the interference source device, the transmission order of real-time calibration commands is adjusted through a priority scheduling algorithm to obtain an optimized command sequence; Based on the optimized instruction sequence, and by updating the power allocation parameters of the parallel devices through an incremental data fusion method, a real-time power allocation scheme is obtained. The steps for obtaining a real-time power allocation scheme based on an optimized instruction sequence and by updating the power allocation parameters of parallel devices using an incremental data fusion method include: The system acquires real-time operating data of parallel devices, and performs noise reduction and standardization on the collected data through a data preprocessing module to obtain a standardized operating dataset. Based on the standardized operational dataset, an incremental data fusion method is used to update the power allocation parameters and obtain a preliminary power allocation scheme by combining equipment operating efficiency and real-time requirements. Based on the system resource usage of the initial power allocation scheme, if the resource utilization rate exceeds the preset threshold, the power allocation parameters are adjusted to obtain a resource-optimized power allocation scheme. The power allocation balance of the resource optimization power allocation scheme is classified and evaluated based on the support vector machine algorithm to determine whether the power allocation scheme meets the balance requirement. If the power allocation scheme meets the balance requirement, time series data of power allocation is generated to obtain the power allocation execution plan. According to the power allocation execution plan, the data encapsulation module encapsulates the allocation parameters into control commands to generate the final real-time power allocation command sequence. Based on the real-time power allocation command sequence, the transmission status of the command sequence is monitored in real time. If the transmission delay exceeds the preset threshold, the transmission path is reallocated and an optimized transmission strategy is determined. Based on the real-time power allocation scheme, and through real-time monitoring of voltage stability data, it is determined whether there is an abnormal voltage fluctuation, and the execution effect of the calibration command is obtained; if the voltage fluctuation is abnormal, the multi-source sensor data is segmented through dynamic time window analysis technology to obtain local energy consumption deviation characteristics. Based on the characteristics of local energy consumption deviation, the energy consumption data is smoothed by the Kalman filter algorithm to obtain stable energy consumption evaluation results under high-frequency production cycle time. Based on the stable energy consumption assessment results, distributed data aggregation technology is used to perform hierarchical fusion of equipment-level and factory-level data to obtain a consistent energy efficiency assessment model.

2. The energy consumption detection method based on the Industrial Internet of Things according to claim 1, characterized in that: The steps for time-synchronized data acquisition, based on timestamp technology, of power curves, current waveforms, and voltage stability data collected from multiple sensors, include: Power curves, current waveforms, and voltage stability data are acquired from multi-source sensors using timestamp technology to generate an initial time series dataset. The timestamps in the initial time series dataset are aligned using a linear interpolation algorithm to obtain a time-synchronized intermediate dataset. If the timestamp deviation of the intermediate dataset exceeds a preset threshold, the power curve, current waveform, and voltage stability data are adjusted using a sliding window method to generate a deviation-corrected dataset. Based on the bias-corrected dataset, time-series features of power curves, current waveforms, and voltage stability are extracted to obtain a feature-enhanced dataset. The feature enhancement dataset is processed using the Kalman filter algorithm to eliminate noise interference and obtain a smoothed dataset. Based on the smoothed dataset, a dynamic time warping algorithm is used to optimize the temporal differences of multi-source data to obtain a high-precision synchronized dataset. Based on the high-precision synchronization dataset, the final time-series dataset is generated, and it is determined whether the synchronization accuracy meets the preset threshold requirements.

3. The energy consumption detection method based on the Industrial Internet of Things according to claim 1, characterized in that: The steps for analyzing the frequency domain characteristics of the current waveform in the time-series synchronization dataset using Fast Fourier Transform to determine whether there are abnormal frequency components caused by electromagnetic interference and to obtain the interference source device include: Based on the time-series data acquisition system, the raw current waveform signal is acquired, and the baseline drift and high-frequency noise are removed by the preprocessing module to obtain standardized current waveform data. The standardized current waveform data is converted into the frequency domain based on the Fast Fourier Transform algorithm to obtain the spectral amplitude distribution, phase information, and energy density of each frequency component. A frequency reference template is established based on the spectral amplitude distribution. If the detected frequency component exceeds the preset threshold range, it is marked as an abnormal frequency point, and a set of abnormal frequency components is obtained. Analyze the harmonic characteristics of each frequency point in the abnormal frequency component set, determine the interference type attribute, and determine the electromagnetic interference mode classification result; Based on the electromagnetic interference pattern classification results matched with the device frequency fingerprint database, if the interference frequency matches the device characteristic frequency, the type of potential interference source device is determined. Acquire synchronous collection timestamp information and operating status data of potential interference source devices, verify the time of interference occurrence through time correlation analysis, and obtain the final interference source device location result; A frequency domain interference characteristic report is generated based on the location results of the interference source device, recording abnormal frequency parameters and corresponding device identification information.

4. The energy consumption detection method based on the Industrial Internet of Things according to claim 1, characterized in that: The steps for obtaining an optimized command sequence based on the interference source device and by adjusting the transmission order of real-time calibration commands using a priority scheduling algorithm include: Based on the signal strength and frequency range parameters of interference source devices in the industrial Internet of Things system, and by analyzing the interference characteristics of each device type through the signal analysis module, an equipment interference impact assessment matrix is ​​obtained. Based on the interference intensity values ​​in the equipment interference impact assessment matrix, and through a weighted scoring mechanism, the priority weight of each calibration command is calculated. If the interference intensity exceeds the preset threshold, the priority weight of the corresponding command is increased to obtain the initial priority ranking table. The initial priority sorting table is processed based on the priority scheduling algorithm. Combining real-time requirements and response time constraints, if the instruction response time is less than the system's preset time window, the current sorting position is maintained to obtain the final instruction scheduling order. Based on the current available system resources, if the resource utilization rate exceeds the system's capacity limit, the instruction execution time interval is adjusted to obtain a resource optimization configuration scheme. The system acquires real-time status information and bandwidth utilization data of the transmission channels, distributes calibration commands to different transmission channels through a load balancing algorithm, and reallocates command transmission paths and determines channel allocation strategies if the load of a channel exceeds the capacity threshold. Based on the channel allocation strategy and instruction scheduling order, a timing controller is used to generate accurate instruction sending timestamps. Combined with the estimated execution time of each instruction, the time conflict between instructions is determined, and a time scheduling matrix is ​​obtained. According to the timing arrangement in the time scheduling matrix, the optimized calibration instructions are encapsulated in the transmission order by the instruction encapsulation module. If the instruction data packet size exceeds the transmission unit limit, it is fragmented to generate the final optimized instruction sequence.

5. The energy consumption detection method based on the Industrial Internet of Things according to claim 1, characterized in that: Based on a real-time power allocation scheme and through real-time monitoring of voltage stability data, the steps to determine whether there are abnormal voltage fluctuations and to obtain the execution effect of calibration commands include: Based on voltage stability data, voltage fluctuation characteristics are obtained through time series analysis. If the voltage fluctuation characteristics exceed a preset threshold, an anomaly detection algorithm is used to determine whether an anomaly exists and to identify the abnormal state. Based on the abnormal state, a pre-established calibration model is used to generate calibration instructions, which are then transmitted to the power distribution unit to obtain execution feedback. Based on the execution feedback, the power allocation status is extracted from the execution feedback, and the execution effect of the calibration command is judged by the data comparison method. If the execution effect does not meet the preset standard, the calibration command is adjusted according to the power allocation status to obtain the optimized command. The power distribution unit is calibrated a second time according to the optimization instructions to obtain the final execution effect.

6. The energy consumption detection method based on the Industrial Internet of Things according to claim 1, characterized in that: Based on the characteristics of local energy consumption deviations, the steps for smoothing energy consumption data using the Kalman filter algorithm to obtain stable energy consumption evaluation results under high-frequency production cycles include: Obtain raw energy consumption data collected during the production process, extract local energy consumption deviation features, and obtain an initial energy consumption deviation dataset; The initial energy consumption deviation dataset is smoothed using the Kalman filter algorithm to obtain smoothed energy consumption data. Based on smoothed energy consumption data, the energy consumption change trend under high-frequency production cycle is calculated to obtain an energy consumption trend sequence; if the fluctuation amplitude of the energy consumption trend sequence exceeds a preset threshold, the outlier is marked to obtain the marked energy consumption trend sequence. Based on time window analysis of the labeled energy consumption trend series, stable energy consumption data is extracted to obtain stable energy consumption evaluation results; Based on the stable energy consumption assessment results and combined with the production cycle frequency, the energy consumption assessment accuracy is calculated to obtain the accuracy assessment value; The production process monitoring system is updated in real time based on the accuracy evaluation values ​​to generate optimized monitoring parameters.

7. An energy consumption detection system based on the Industrial Internet of Things, characterized in that, include: The timing alignment module is used to perform timing alignment on the power curves, current waveforms and voltage stability data collected by multi-source sensors according to the timestamp marking technology, and obtain a timing synchronization dataset; The frequency domain analysis module is used to analyze the frequency domain characteristics of the current waveform in the time-series synchronization dataset based on the Fast Fourier Transform, determine whether there are abnormal frequency components caused by electromagnetic interference, and obtain the interference source device. The scheduling optimization module is used to adjust the transmission order of real-time calibration commands based on the interference source device and through a priority scheduling algorithm to obtain an optimized command sequence. The data fusion module is used to update the power allocation parameters of parallel devices according to the optimized instruction sequence and through incremental data fusion method to obtain a real-time power allocation scheme; The steps for obtaining a real-time power allocation scheme based on an optimized instruction sequence and by updating the power allocation parameters of parallel devices using an incremental data fusion method include: The system acquires real-time operating data of parallel devices, and performs noise reduction and standardization on the collected data through a data preprocessing module to obtain a standardized operating dataset. Based on the standardized operational dataset, an incremental data fusion method is used to update the power allocation parameters and obtain a preliminary power allocation scheme by combining equipment operating efficiency and real-time requirements. Based on the system resource usage of the initial power allocation scheme, if the resource utilization rate exceeds the preset threshold, the power allocation parameters are adjusted to obtain a resource-optimized power allocation scheme. The power allocation balance of the resource optimization power allocation scheme is classified and evaluated based on the support vector machine algorithm to determine whether the power allocation scheme meets the balance requirement. If the power allocation scheme meets the balance requirement, time series data of power allocation is generated to obtain the power allocation execution plan. According to the power allocation execution plan, the data encapsulation module encapsulates the allocation parameters into control commands to generate the final real-time power allocation command sequence. Based on the real-time power allocation command sequence, the transmission status of the command sequence is monitored in real time. If the transmission delay exceeds the preset threshold, the transmission path is reallocated and an optimized transmission strategy is determined. The voltage monitoring module is used to determine whether there are abnormal voltage fluctuations based on the real-time power distribution scheme and by monitoring the voltage stability data in real time, so as to obtain the execution effect of the calibration command. The segmented processing module is used to determine whether the voltage fluctuation is abnormal. If an abnormality is found, the multi-source sensor data is segmented using dynamic time window analysis technology to obtain local energy consumption deviation characteristics. The filtering module is used to smooth the energy consumption data based on the local energy consumption deviation characteristics using the Kalman filtering algorithm, so as to obtain the stable energy consumption evaluation results under high-frequency production cycle time. The hierarchical aggregation module is used to perform hierarchical fusion of equipment-level and factory-level data based on the stable energy consumption evaluation results, using distributed data aggregation technology to obtain a consistent energy efficiency evaluation model.

8. An electronic device, characterized in that, include: A memory on which computer programs are stored; A processor for executing a computer program in memory to implement the energy consumption detection method based on the Industrial Internet of Things as described in any one of claims 1 to 7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the program implements the energy consumption detection method based on the Industrial Internet of Things as described in any one of claims 1 to 7.

Images