Multi-parameter cooperative detection high-precision instrument data processing method and system

Abstract

The embodiment of the invention discloses a multi-parameter cooperative detection high-precision instrument data processing method and system, and relates to the technical field of instruments and data processing. The method comprises the following steps: synchronously acquiring measured parameters and environmental interference parameters of target equipment through a multi-source sensor array to form a time sequence data set; a dynamic coupling analysis model is constructed, and the time series data set is converted into a space incidence matrix by using a Grubrum angle field; establishing a variable weight collaborative decision model based on a long short-term memory network, and dynamically adjusting the contribution weight of the acquired data in a final detection result according to a real-time working condition; carrying out online calibration on the output of the sensor by adopting a hybrid calibration algorithm; calculating the compensation amount of the environmental interference parameter through a Bayesian network; and outputting high-precision detection data. According to the method, the time domain parameters can be mapped into the two-dimensional spatial features in real time, so that the recognition precision of the pressure-temperature coupling coefficient is improved, and the multi-parameter collaborative decision error is reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of instruments and data processing technology, in particular to a high-precision instrument data processing method and system for multi-parameter collaborative detection. BACKGROUND

[0002] In the field of electric power, accurate instrument data processing is crucial to ensure equipment reliability and measurement accuracy. Currently, high-precision electric power instruments and meters such as single three-phase multifunctional standard electric energy meters and single three-phase field verification instruments are widely used in electric power systems. Due to their high measurement accuracy and wide range, the data processing process is complex.

[0003] Common electronic instrument data processing methods include data acquisition, data transmission, data storage, and data analysis. Through data processing, we can better understand the data generated during the experiment, draw useful conclusions, and make reasonable decisions. Therefore, it is necessary to master common electronic instrument data processing methods.

[0004] The existing instrument data processing method has the following defects: 1. Single parameter independent processing: traditional instruments use separate processing for multiple parameters, ignoring the physical coupling relationship between parameters; 2. Static compensation model: existing compensation algorithms are based on fixed correction coefficients, which cannot adapt to dynamic changes in working conditions; 3. Error accumulation: when multiple sensor data fusion is performed, an error transmission suppression mechanism is not established, resulting in a decrease in system-level accuracy. SUMMARY

[0005] Therefore, the present application provides a high-precision instrument data processing method and system for multi-parameter collaborative detection to solve the problems of ignoring the physical coupling relationship between parameters, failing to adapt to dynamic changes in working conditions, and error accumulation leading to a decrease in system-level accuracy.

[0006] In one aspect, the present application provides a high-precision instrument data processing method for multi-parameter collaborative detection, comprising: Step S1: Synchronously collecting N measured parameters and M environmental interference parameters of a target device through a multi-source sensor array to form a time series data set; Step S2: Building a dynamic coupling analysis model, converting the time series data set into a spatial correlation matrix using a Gram angle field, and quantifying the non-linear coupling strength between parameters; Step S3: Building a variable weight collaborative decision-making model based on a long short-term memory network, and dynamically adjusting the contribution weight of the data collected in step S1 in the final detection result according to the real-time working condition; Step S4: using a hybrid calibration algorithm to calibrate the measured parameters of the processed sensor final detection output online, wherein the hybrid calibration algorithm combines state estimation of Kalman filtering and parameter optimization of particle swarm optimization; Step S5: calculating the compensation amount of the environmental interference parameter through the Bayesian network to correct the output value of the corresponding sensor, and feeding back the compensation result to the variable weight collaborative decision model; Step S6: outputting the high-precision detection data obtained through the above step S5.

[0007] Further, the step S2 comprises: Step S2.1: normalizing the time series of each parameter to eliminate the dimensional difference; Step S2.2: mapping the one-dimensional time series into a two-dimensional matrix through the Gram angle field to retain the time domain correlation; Step S2.3: extracting the cross-parameter coupling features in the matrix through a convolutional neural network to generate an N×N correlation matrix.

[0008] Further, in the step S2.3, the update frequency of the correlation matrix is positively correlated with the parameter change rate, and specifically meets the following condition: ; Wherein, α is a preset sensitivity coefficient, x_i is the instantaneous value of the i-th parameter, and the vertical line represents a conditional separator.

[0009] Further, the step S3 comprises: Step S3.1: establishing a bidirectional LSTM network containing an attention mechanism; Step S3.2: setting a forgetting factor γ=0.9-0.99 for dynamically attenuating the weight influence of historical data; Step S3.3: outputting the real-time weight distribution of each parameter through a Softmax function.

[0010] Further, the step S4 comprises: Step S4.1: using Kalman filtering to eliminate the first-order error of sensor noise; Step S4.2: using an improved particle swarm optimization algorithm to compensate for the second-order nonlinear error.

[0011] Further, in the step S4.2, the inertia weight ω is adaptively adjusted according to the formula ω=ω_max-(ω_max-ω_min)·(k / K) 2 , k is the current iteration number, K is the total iteration number, ω_max is the maximum value of the inertia weight, and ω_min is the minimum value of the inertia weight.

[0012] Further, the step S5 comprises: Step S5.1: constructing a directed acyclic graph of environmental interference parameters and measurement errors, and obtaining node conditional probabilities through historical data training; Step S5.2: performing posterior probability inference by using a Markov chain Monte Carlo method, and outputting an optimal compensation amount.

[0013] Further, the step S1 comprises: applying a dynamic time warping algorithm to the sensor data to solve the timing misalignment problem caused by asynchronous sampling, and achieving an alignment accuracy of 1 / 100 of the sampling period.

[0014] In another aspect, the application provides a high-precision instrument and meter data processing system for multi-parameter collaborative detection, comprising: a multi-parameter synchronous acquisition module, configured to synchronously acquire N measured parameters and M environmental interference parameters of a target device through a multi-source sensor array, and form a time series data set; a coupling analysis module, configured to construct a dynamic coupling analysis model, convert the time series data set into a spatial correlation matrix by using a Gram angle field, and quantify the nonlinear coupling strength between parameters; a dynamic decision module, configured to establish a variable weight collaborative decision model based on a long short-term memory network, and dynamically adjust the contribution weight of the data acquired in the multi-parameter synchronous acquisition module in the final detection result according to a real-time working condition; an online calibration module, configured to perform online calibration on the measured parameters in the final detection output of the processed sensor by using a hybrid calibration algorithm, wherein the hybrid calibration algorithm combines state estimation of Kalman filtering and parameter optimization of particle swarm optimization; an environmental compensation module, configured to calculate a compensation amount of the environmental interference parameters by using a Bayesian network to correct the output value of the corresponding sensor, and feed back the compensation result to the variable weight collaborative decision model; a data output module, configured to output high-precision detection data processed by the environmental compensation module.

[0015] Further, the dynamic decision module adopts a pipeline architecture, comprising: a first-stage pipeline: parameter weight calculation, with a period of 5 ns; a second-stage pipeline: environmental compensation amount superposition, with a period of 3 ns; a third-stage pipeline: data validity verification, with a period of 2 ns.

[0016] The application has the following beneficial effects: The multi-parameter cooperative detection high-precision instrument and meter data processing method and system of the application, by setting the correlation matrix of the Gram angle field (GAF) combined with the convolutional neural network (CNN), maps the time domain parameters into two-dimensional space features in real time, improves the pressure-temperature coupling coefficient identification accuracy, and reduces the multi-parameter cooperative decision error. The long short-term memory network (LSTM) with a forgetting factor (γ=0.95) is used to calculate the parameter weight in real time, reduce the weight adjustment delay when the working condition changes, and deploy the Bayesian network and Markov chain Monte Carlo (MCMC) inference engine to calculate the environmental compensation in real time, reduce the pressure measurement deviation caused by the steam density fluctuation, and improve the signal-to-noise ratio (SNR) under electromagnetic interference (EMI). The dynamic time warping (DTW) hardware accelerator is used to align the asynchronous data in real time, reduce the phase difference between the flow and pressure signals, improve the timeliness of data fusion, and realize a three-stage pipeline architecture through the field programmable gate array (FPGA) to execute the LSTM inference in real time, solving the problems of the existing instrument and meter data processing method, such as separate processing of multiple parameters by the instrument, ignoring the physical coupling relationship between parameters, compensating algorithm based on fixed correction coefficient, unable to adapt to dynamic changes in working conditions, and no error transmission suppression mechanism established during multi-sensor data fusion, resulting in a decline in system-level accuracy. BRIEF DESCRIPTION OF DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, a brief introduction will be given below to the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and those skilled in the art can also obtain other drawings according to these drawings without any creative effort.

[0018] Figure 1 The flowchart of the multi-parameter cooperative detection high-precision instrument and meter data processing method of the present application; Figure 2 The structure diagram of the multi-parameter cooperative detection high-precision instrument and meter data processing system of the present application. DETAILED DESCRIPTION

[0019] The embodiments of the present application will be described in detail below with reference to the drawings.

[0020] It should be clear that the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without any creative effort are within the scope of protection of the present application.

[0021] On the one hand, the embodiments of the present application provide a multi-parameter cooperative detection high-precision instrument and meter data processing method, as shown in the following formula (1): Figure 1As shown, comprising the following steps: S1, synchronously collecting N measured parameters of the target device and M environmental interference parameters by a multi-source sensor array, to form a time series data set; In this step, the target device refers to an industrial device, instrument or system that needs to be monitored or controlled, for example: industrial scenarios: pressure pipelines, reaction kettles, power transformers, etc.; other fields: environmental monitoring stations, intelligent instruments, etc. The measured parameters (N) refer to the key physical or chemical quantities that need to be directly measured during the operation of the target device, which are usually related to the core function or state of the device, for example: industrial pressure pipelines: pressure (P), temperature (T), flow (F), etc.; power equipment: voltage, current, power factor, etc. The environmental interference parameters (M) refer to external environmental factors that may affect the measurement accuracy of the measured parameters, which need to be monitored and compensated for their interference, for example: industrial scenarios: environmental temperature, humidity, electromagnetic interference (EMI), steam density, vibration, etc.; general interference: sensor noise, power fluctuations, sampling time deviation, etc. Among them, N and M are integers greater than 1.

[0022] As an optional embodiment, the step S1 comprises: Applying a dynamic time warping (DTW) algorithm to the multi-source sensor data to solve the time sequence misalignment problem caused by asynchronous sampling, and the alignment accuracy reaches 1 / 100 of the sampling period.

[0023] S2, constructing a dynamic coupling analysis model, converting the time series data set into a spatial correlation matrix by Gramian Angular Field (GAF), and quantifying the nonlinear coupling strength between parameters; As an optional embodiment, this step can include: S2.1 Normalizing the time series (data) of each parameter to eliminate dimensional differences; S2.2 Mapping one-dimensional time series (i.e. the time series of each parameter) to a two-dimensional matrix by Gramian Angular Field to preserve time domain correlation; S2.3 Extracting cross-parameter coupling features in the matrix by using a convolutional neural network (CNN) to generate an N x N correlation matrix.

[0024] Preferably, the update frequency of the spatial correlation matrix is positively correlated with the parameter change rate, which satisfies the following conditions: Wherein, a is a preset sensitivity coefficient, x_i is the instantaneous value of the i-th parameter, and the vertical bar represents a conditional separator.

[0025] S3, a variable weight collaborative decision-making model is established based on a long short term memory (LSTM) network, and the contribution weight of the data collected in the step S1 in the final detection result is dynamically adjusted according to a real-time working condition; As an optional embodiment, the present step can comprise: S3.1, a bidirectional LSTM network containing an attention mechanism is established; Specifically, the real-time weight distribution output by the bidirectional LSTM network is directly used for subsequent multi-parameter data fusion (step S5) through Softmax normalization, and determines the contribution proportion of each parameter in the final detection result. The environmental interference compensation amount calculated by the Bayesian network in step S5 is fed back to the Bi-LSTM model (step S3) to dynamically adjust the weight to offset the influence of environmental interference. The optimized sensor data of the hybrid calibration algorithm in step S4 is used as the input of the LSTM, which improves the accuracy of the weight calculation.

[0026] S3.2, a forgetting factor γ=0.9-0.99 is set to dynamically attenuate the weight influence of historical data; S3.3, the real-time weight distribution of each parameter is output through a Softmax function.

[0027] In specific implementation, the parameter importance can be used as a training label to make the LSTM learn to output the original score related to the parameter importance through supervised learning. The Softmax feature ensures that the output value is in the interval (0, 1) and the sum is 1, which meets the basic requirements of weight allocation. Dynamic adaptability: LSTM can dynamically adjust the output score according to the changes of input data, so as to reflect the changes of the relative importance of each parameter under different working conditions.

[0028] In the present step S3.3, the sum of the real-time weights of each parameter is 1, that is, Σw_i=1, where w_i is the real-time weight of the i-th parameter.

[0029] Specifically, the specific working principle of LSTM is as follows: (1) Network structure Bidirectional design: containing forward and backward LSTM layers, respectively scanning time series data from past to future and future to past, fully capturing context dependence.

[0030] Forward LSTM: processing historical data; Backward LSTM: processing reverse data.

[0031] Attention mechanism: assigning importance scores to different time steps to highlight key periods (such as moments of temperature mutation).

[0032] (2) Dynamic weight calculation Forget factor (gamma = 0.9-0.99): control the decay rate of historical memory, for example, gamma = 0.95 means to retain 95% of the historical information, adapt to gradual and sudden changes.

[0033] Softmax output: map the LSTM hidden state to a weight vector (w) that satisfies , ensuring weight normalization.

[0034] (3) Synergy with physical coupling Input features: receive the spatial correlation matrix (quantitative parameter nonlinear coupling strength) from the GAF-CNN as an auxiliary basis for weight decision-making.

[0035] S4, the hybrid calibration algorithm is used to calibrate the measured parameters of the final detection output of the processed sensor online, and the hybrid calibration algorithm combines the state estimation of Kalman filter and the parameter optimization of particle swarm optimization (PSO); As an optional embodiment, the step can include: S4.1 First-order error elimination of sensor noise is performed using Kalman filtering; S4.2 Improved particle swarm optimization algorithm is used for second-order nonlinear error compensation.

[0036] In this step S4.2, preferably, the inertia weight ω is adaptively adjusted according to the formula ω = ω_max-(ω_max-ω_min)·(k / K) 2 , where k is the current iteration number, K is the total iteration number, ω_max is the maximum value of inertia weight, and ω_min is the minimum value of inertia weight.

[0037] S5, the compensation amount of the environmental interference parameter is calculated through the Bayesian network to correct the output value of the corresponding sensor, and the compensation result is fed back to the variable weight collaborative decision-making model; In this step, the following processing is performed synchronously: 1) directly correct the output value of the corresponding sensor; 2) as a feature input to the variable weight collaborative decision-making model, dynamically adjust the weight distribution coefficient of each parameter.

[0038] As an optional embodiment, the step can include: S5.1 Build a directed acyclic graph (DAG) of environmental interference parameters and measurement errors, and the node conditional probability is obtained by training historical data; ​S5.2 Markov Chain Monte Carlo (MCMC) method is used for posterior probability inference, and the optimal compensation amount is output.

[0039] In this step S5.2, preferably, the optimal compensation amount ΔY = argmax P(Y|X_E), wherein X_E is the environmental parameter vector, argmax is a function, P represents the conditional probability, and ΔY is the compensation amount for the measured parameter Y.

[0040] In the above steps S2-S5, in the specific implementation, the nonlinear coupling strength: in the dynamic coupling modeling, the time series data (such as pressure, temperature) is converted into a two-dimensional matrix through GAF, and the time domain correlation is retained. The CNN extracts cross-parameter features to generate a coupling coefficient matrix. The LSTM network analyzes the parameter time sequence relationship and outputs the weight combined with the coupling matrix. The Kalman filter eliminates the first-order noise of the sensor, the improved PSO algorithm compensates the second-order nonlinear error, the Bayesian network compensates the input environmental parameters X_E (such as temperature, steam density), and the optimal compensation amount ΔY = argmax P(Y|X_E) is output. The synergistic workflow of the three, taking industrial pressure pipeline monitoring as an example: Data acquisition: synchronously measure pressure (P), temperature (T), flow (F), and environmental temperature and steam density.

[0041] Coupling analysis: through GAF+CNN, it is found that the coupling coefficient of P and T is 0.72.

[0042] Weight distribution: when the temperature suddenly changes, the weight of T is increased from 0.3 to 0.4, and the weight of P is reduced to 0.5.

[0043] Compensation calculation: according to the change of steam density, the Bayesian network outputs the pressure compensation amount ΔP = 0.23 MPa.

[0044] Final output: combine the weight and the compensation amount to output the corrected high-precision pressure value.

[0045] S6, output the high-precision detection data obtained by the step S5.

[0046] Compared with the existing technology, the present invention provides a high-precision instrumentation data processing method for multi-parameter collaborative detection, which has the advantages of improving data processing accuracy, reducing response speed, and having anti-interference capabilities. It solves the problems of existing instrumentation data processing methods, such as the instrument using discrete processing for multiple parameters, ignoring the physical coupling relationship between parameters, the compensation algorithm is mostly based on fixed correction coefficients and cannot adapt to dynamic changes in working conditions, and the error transmission suppression mechanism is not established when multi-sensor data is fused, resulting in a decrease in system-level accuracy. The technical solution of the present invention has the following beneficial effects: 1. This high-precision instrument data processing method for multi-parameter collaborative detection maps time-domain parameters into two-dimensional spatial features in real time by setting a correlation matrix that combines the Gram Angular Field (GAF) with CNN, thereby improving the accuracy of pressure-temperature coupling coefficient identification and reducing multi-parameter collaborative decision-making errors.

[0047] 2. This high-precision instrumentation data processing method for multi-parameter collaborative detection uses an LSTM network with a forgetting factor (γ=0.95) to calculate parameter weights in real time, reducing the delay in weight adjustment when operating conditions suddenly change. By deploying a Bayesian network and MCMC inference engine, environmental compensation is calculated in real time, reducing pressure measurement deviations caused by steam density fluctuations and improving the signal-to-noise ratio (SNR) under electromagnetic interference (EMI).

[0048] 3. This high-precision instrumentation data processing method for multi-parameter collaborative detection uses a dynamic time warping (DTW) hardware accelerator to align asynchronous data in real time, reduce the phase difference between flow and pressure signals, and improve the timeliness of data fusion. It implements a three-stage pipeline architecture through an FPGA (field programmable gate array) and performs LSTM inference in real time.

[0049] Implementation Case: Multi-parameter Monitoring System for Industrial Pressure Pipelines 1. Technical Background A waste incineration power plant in Zhejiang Province needs to monitor the pressure, temperature, and flow parameters of high-pressure pipelines in real time. Traditional instruments have an error of ±2.5% under steam disturbance conditions, which cannot meet production safety requirements.

[0050] 2. Implementation steps 2.1 Data Collection Sensor configuration: piezoelectric pressure sensor (range 0-10MPa, accuracy 0.1%), PT100 temperature sensor (-50~300℃, ±0.5℃), vortex flowmeter (accuracy level 0.5).

[0051] Synchronous sampling: IEEE 1588 protocol is used to achieve three-channel time synchronization (deviation < 1μs).

[0052] 2.2 Dynamic coupling modeling Convert the time series data (sampling rate 1 kHz) of pressure (P), temperature (T), and flow rate (F) into a Gram matrix: # Example code snippet from pyts.image import GramianAngularField gaf = GramianAngularField(image_size=100) X_gaf = gaf.fit_transform(X.T) # X is the normalized data of P / T / F Extract features through dynamic coupling analysis model to obtain correlation matrix: 2.3 Adaptive compensation Dynamic adjustment of LSTM weights (hidden layer 128 nodes): Steady-state weights: P(0.6) / T(0.3) / F(0.1).

[0053] When temperature suddenly changes (ΔT>5℃ / s): automatically adjust to P(0.5) / T(0.4) / F(0.1).

[0054] Environmental compensation: calculate the pressure measurement deviation caused by the change of steam density through Bayesian network, compensation amount ΔP=0.23MPa (when the steam density changes from 2.1 to 3.5kg / m³).

[0055] 3. Performance verification Performance verification is shown in Table 1: Table 1 On the other hand, the embodiments of the present application provide a high-precision instrument and meter data processing system for multi-parameter collaborative detection, as shown in Figure 2 , comprising: A multi-parameter synchronous acquisition module 10 is used to synchronously acquire N measured parameters and M environmental interference parameters of a target device through a multi-source sensor array to form a time series data set. In specific implementation, the module can include a high-precision ADC circuit and an anti-aliasing filter. A coupling analysis module 20 is used to construct a dynamic coupling analysis model, convert the time series data set into a spatial correlation matrix using a Gram angular field, and quantify the nonlinear coupling strength between parameters. In specific implementation, the module can be configured to perform GAF transformation and CNN feature extraction. The dynamic decision module 30 is configured to establish a variable weight collaborative decision model based on a long short-term memory network, and dynamically adjust the contribution weight of the data collected by the multi-parameter synchronous acquisition module 10 in the final detection result according to real-time working conditions. The online calibration module 40 is configured to perform online calibration on the measured parameters in the final detection output of the processed sensor by using a hybrid calibration algorithm combining state estimation of Kalman filtering and parameter optimization of particle swarm optimization. The environmental compensation module 50 is configured to calculate a compensation amount of an environmental interference parameter by using a Bayesian network to correct the output value of the corresponding sensor, and feed back the compensation result to the variable weight collaborative decision model. The data output module 60 is configured to output high-precision detection data processed by the environmental compensation module 50.

[0056] The device of the embodiment can be used to perform the method of the embodiment. Figure 1 The technical solutions of the method embodiment are similar in implementation principle and technical effects, and will not be described here.

[0057] Preferably, the dynamic decision module 60 adopts a pipeline architecture, which includes: The first stage pipeline: parameter weight calculation, cycle 5ns; The second stage pipeline: environmental compensation amount superposition, cycle 3ns; The third stage pipeline: data validity verification, cycle 2ns.

[0058] The embodiment of the application further provides a computer readable storage medium, and the computer readable storage medium stores a computer program.

[0059] The embodiment of the application further provides an application program, and the application program is executed to implement the method provided by any method embodiment of the application.

[0060] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply the existence of any such actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or device comprising the element.

[0061] Each embodiment in this specification is described in a related manner. The same or similar parts between the embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments. In particular, for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple. For the relevant parts, refer to the partial description of the method embodiment. For the convenience of description, the above device is described by dividing it into various units / modules according to their functions. Of course, when implementing the present invention, the functions of each unit / module can be implemented in the same or multiple software and / or hardware.

[0062] Those skilled in the art will appreciate that all or part of the processes in the above-described method embodiments can be implemented by instructing the relevant hardware through a computer program. The program can be stored in a computer-readable storage medium, and when executed, the program can include the processes in the above-described method embodiments. The storage medium can be a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

[0063] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included in the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. A high-precision instrumentation data processing method for multi-parameter collaborative detection, characterized in that: include: Step S1: synchronously collect N measured parameters of the target device and M environmental interference parameters through a multi-source sensor array to form a time series data set; Step S2: constructing a dynamic coupling analysis model, converting the time series data set into a spatial correlation matrix using the Gram angle field, and quantifying the nonlinear coupling strength between various parameters; Step S3: establishing a variable weight collaborative decision-making model based on the long short-term memory network, and dynamically adjusting the contribution weight of the data collected in step S1 in the final detection result according to the real-time working conditions; Step S4: online calibration of the measured parameters of the final detection output of the processed sensor is performed using a hybrid calibration algorithm, wherein the hybrid calibration algorithm combines state estimation of Kalman filtering with parameter optimization of particle swarm optimization; Step S5: Calculating the compensation amount of the environmental interference parameter through the Bayesian network to correct the output value of the corresponding sensor, and feeding back the compensation result to the variable weight collaborative decision model; Step S6: Output the high-precision detection data obtained through the processing in step S5.

2. The high-precision instrumentation data processing method for multi-parameter collaborative detection according to claim 1 is characterized in that: The step S2 of constructing the dynamic coupling analysis model specifically includes: Step S2.1: Normalize the time series of each parameter to eliminate dimensional differences; Step S2.2: Map the one-dimensional time series into a two-dimensional matrix using the Gram angle field to preserve the time domain correlation; Step S2.3: Use a convolutional neural network to extract cross-parameter coupling features in the matrix and generate an N×N correlation matrix.

3. The high-precision instrumentation data processing method for multi-parameter collaborative detection according to claim 2 is characterized in that: In step S2.3, the update frequency of the correlation matrix is ​​positively correlated with the parameter change rate, specifically satisfying the following conditions: ； Where α is the preset sensitivity coefficient, x_i is the instantaneous value of the i-th parameter, and the vertical line represents the conditional separator.

4. The high-precision instrumentation data processing method for multi-parameter collaborative detection according to claim 1 is characterized in that: The step S3 comprises: Step S3.1: Build a bidirectional LSTM network with attention mechanism. Step S3.2: Set the forgetting factor γ to 0.9-0.99 to dynamically attenuate the weight influence of historical data; Step S3.3: Output the real-time weight distribution of each parameter through the Softmax function.

5. The high-precision instrumentation data processing method for multi-parameter collaborative detection according to claim 1 is characterized in that: The step S4 comprises: Step S4.1: Use Kalman filtering to eliminate the first-order error of sensor noise; Step S4.2: Use the improved particle swarm optimization algorithm to perform second-order nonlinear error compensation.

6. The high-precision instrumentation data processing method for multi-parameter collaborative detection according to claim 5 is characterized in that: In step S4.2, the inertia weight ω is calculated according to the formula ω=ω_max-(ω_max-ω_min)·(k / K) 2 Adaptive adjustment, k is the current iteration number, K is the total iteration number, ω_max is the maximum inertia weight, and ω_min is the minimum inertia weight.

7. The high-precision instrumentation data processing method for multi-parameter collaborative detection according to claim 1 is characterized in that: The step S5 comprises: Step S5.1: Construct a directed acyclic graph of environmental interference parameters and measurement errors, and the node conditional probabilities are obtained through historical data training; Step S5.2: Use the Markov Chain Monte Carlo method to perform posterior probability inference and output the optimal compensation amount.

8. The high-precision instrumentation data processing method for multi-parameter collaborative detection according to any one of claims 1 to 7, characterized in that: The step S1 comprises: A dynamic time warping algorithm is applied to multi-source sensor data to solve the timing misalignment problem caused by asynchronous sampling, and the alignment accuracy reaches 1 / 100 of the sampling period.

9. A high-precision instrumentation data processing system for multi-parameter collaborative detection, characterized in that: include: Multi-parameter synchronous acquisition module, used to synchronously acquire N measured parameters of the target device and M environmental interference parameters through a multi-source sensor array to form a time series data set; A coupling analysis module is used to construct a dynamic coupling analysis model, convert the time series data set into a spatial correlation matrix using the Gram angle field, and quantify the nonlinear coupling strength between various parameters; A dynamic decision-making module is used to establish a variable weight collaborative decision-making model based on a long short-term memory network, and dynamically adjust the contribution weight of the data collected in the multi-parameter synchronous acquisition module in the final detection result according to the real-time working conditions; An online calibration module is used to calibrate the measured parameters of the final detection output of the processed sensor online using a hybrid calibration algorithm that combines state estimation using Kalman filtering with parameter optimization using particle swarm optimization; An environmental compensation module, configured to calculate the compensation amount of the environmental interference parameter through a Bayesian network to correct the output value of the corresponding sensor, and feed the compensation result back to the variable weight collaborative decision-making model; The data output module is used to output the high-precision detection data processed by the environmental compensation module.

10. The high-precision instrumentation data processing system for multi-parameter collaborative detection according to claim 9 is characterized in that: The dynamic decision module adopts a pipeline architecture, including: First-stage pipeline: parameter weight calculation, cycle 5ns; Second-stage pipeline: environmental compensation amount superposition, cycle 3ns; Third-level pipeline: data validity check, cycle 2ns.

Images