Industrial park data integration and analysis system based on cloud computing
Through cloud computing technology, the equipment status monitoring signals and operating environment parameter signals in industrial parks are analyzed, and local signal segments are divided for signal decomposition, which solves the problem of interference from monitoring signals and improves the accuracy and quality of data analysis.
Patent Information
- Application Number
- CN202410674753.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-05-29
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2044-05-29
AI Technical Summary
In the prior art, the monitoring signals in industrial parks are affected by complex interference, and the data signal processing and analysis capabilities are weak, resulting in low data quality and large errors, which affects subsequent monitoring and analysis.
Adopt a cloud-based industrial park data integration and analysis system, and through the data acquisition module, interference impact analysis module, local interference analysis module and enhanced monitoring module, the equipment status monitoring signal and operating environment parameter signals are obtained, and the signal responsiveness and responsiveness of the signal are analyzed. Interference degree, divide local signal segments for signal decomposition, calculate local interference confidence, determine interference index, and perform local adaptive de-interference processing.
It improves the accuracy of analysis of disturbed status monitoring signals, obtains better monitoring signals, and enhances the accuracy and quality of data analysis.
Smart Images

Figure CN118584865B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of big data processing technology, and in particular to an industrial park data integration and analysis system based on cloud computing. Background Art
[0002] Cloud computing is a distributed computing network technology that integrates, stores and analyzes a large amount of data generated by various industrial equipment in the park, thereby providing real-time, comprehensive data visualization and decision support, and improving the management efficiency and decision-making ability of the industrial park. With the increasing digitalization and intelligence of industrial production, the demand for data processing in industrial parks is also growing, and the quality of data is extremely important for the monitoring and analysis of industrial park operations.
[0003] In industrial parks, a large number of sensors are deployed to monitor various state parameters of industrial production to ensure the normal operation of system equipment. However, since sensors are affected by electromagnetic fields or environmental changes during the process of collecting industrial data, complex interference fluctuations are superimposed on the collected data signals. Since different equipment in industrial parks are affected by different operating conditions, the commonly used data signal processing is too unified for the processing and analysis of equipment monitoring data that is subject to complex interference conditions in time periods. The correction and enhancement capabilities of monitoring signals subject to complex interference are weak, and the quality of the enhanced data is not high. There may still be large errors, which affect subsequent monitoring and analysis. Summary of the invention
[0004] In order to solve the technical problems in the prior art that the correction and enhancement capability of monitoring signals subject to complex interference is weak, the quality of enhanced data is not high, and there may still be large errors, the purpose of the present invention is to provide an industrial park data integration and analysis system based on cloud computing, and the technical solutions adopted are as follows:
[0005] The present invention provides an industrial park data integration and analysis system based on cloud computing, the system comprising:
[0006] The data acquisition module is used to obtain the status monitoring signal of the equipment and the parameter monitoring signal of each operating environment during the preset time period of each equipment in the industrial park;
[0007] The interference impact analysis module is used to obtain the responsiveness of the state monitoring signal to each parameter monitoring signal according to the degree of correlation between the state monitoring signal of the equipment and the parameter monitoring signal of each operating environment, and the correlation between each operating environment and other operating environments in the parameter monitoring signal; obtain the interference degree of the state monitoring signal according to the responsiveness of the state monitoring signal of the equipment to each parameter monitoring signal, and the stable difference in fluctuation between the state monitoring signal and the parameter monitoring signal of each operating environment;
[0008] A local interference analysis module is used to divide the state monitoring signal of the equipment into at least two local signal segments; obtain the signal segment components of each local signal segment after signal decomposition; obtain the local interference confidence of each local signal segment according to the periodic deviation of the data distribution in the signal segment components between each local signal segment and other local signal segments, and the degree to which each signal segment component in each local signal segment complies with the decomposition rule;
[0009] The enhanced monitoring module is used to determine the interference index of each local signal segment through the interference degree of the state monitoring signal and the local interference confidence of each local signal segment; enhance the state monitoring signal based on the interference index of each local signal segment in the state monitoring signal to obtain the final monitoring signal.
[0010] Furthermore, the method for obtaining the responsiveness of the state monitoring signal to each parameter monitoring signal includes:
[0011] According to the data change correlation between the parameter monitoring signal of each operating environment and the status monitoring signal of the equipment, the change correlation degree of each parameter monitoring signal is obtained;
[0012] According to the change consistency between the parameter monitoring signal of each operating environment and all other parameter monitoring signals, the influence characterization degree of each parameter monitoring signal is obtained;
[0013] The responsiveness of the state monitoring signal to each parameter monitoring signal is determined by the change correlation degree of each parameter monitoring signal and the influence characterization degree of each parameter monitoring signal.
[0014] Furthermore, obtaining the influence characterization degree of each parameter monitoring signal according to the change consistency of the parameter monitoring signal of each operating environment with all other parameter monitoring signals includes:
[0015] Standardize the parameter monitoring signal of each operating environment to obtain a standard curve for each parameter monitoring signal;
[0016] For any parameter monitoring signal, the mean square error between the standard curve of the parameter monitoring signal and the standard curve of each other parameter monitoring signal is taken as the numerical distribution approximation degree between the parameter monitoring signal and each other parameter monitoring signal;
[0017] The standard curve of the parameter monitoring signal and the DTW value of each other parameter monitoring signal are used as the numerical change approximation of the parameter monitoring signal and each other parameter monitoring signal;
[0018] The influence characterization of this parameter monitoring signal is obtained by combining the numerical distribution approximation and numerical change approximation of this parameter monitoring signal with all other parameter monitoring signals. Both the numerical distribution approximation and the numerical change approximation are negatively correlated with the influence characterization.
[0019] Furthermore, the method for obtaining the interference degree includes:
[0020] Determine the abnormality of each parameter monitoring signal according to the difference between the parameter monitoring signal of each operating environment and the corresponding preset parameter expected signal;
[0021] Combining the responsiveness of the state monitoring signal to each parameter monitoring signal and the abnormality of each parameter monitoring signal, the possibility of interference of the state monitoring signal is obtained;
[0022] Determining the difference degree of the state monitoring signal according to the difference between the state monitoring signal and the preset state expected signal;
[0023] The interference degree of the state monitoring signal is obtained according to the difference degree and interference possibility degree of the state monitoring signal.
[0024] Furthermore, the method for obtaining the abnormality degree includes:
[0025] The difference in signal value between each parameter monitoring signal at each moment and the corresponding preset parameter expected signal is taken as the abnormal difference of each parameter monitoring signal at each moment;
[0026] The sum of all abnormal differences is taken as the abnormality of each parameter monitoring signal.
[0027] Furthermore, the method for obtaining the possibility of interference includes:
[0028] For any parameter monitoring signal of an operating environment, the responsiveness of the state monitoring signal to the parameter monitoring signal is multiplied by the abnormality of the parameter monitoring signal to obtain the influence interference degree of the parameter monitoring signal;
[0029] The average values of the impression interference of the parameter monitoring signals of all operating environments are normalized to obtain the interference possibility of the state monitoring signal.
[0030] Furthermore, the method for obtaining the local interference confidence includes:
[0031] determining a principal component among all signal segment components of each local signal segment;
[0032] Determine the period difference of each local signal segment according to the distribution degree of the signal value of the principal component in each local signal segment in the principal components of other local signal segments;
[0033] Determining the decomposition regularity of each local signal segment based on the Gaussianity of each signal segment component in each local signal segment;
[0034] The local interference confidence of each local signal segment is obtained according to the period difference and decomposition regularity of each local signal segment; the period difference and decomposition regularity are both positively correlated with the local interference confidence.
[0035] Furthermore, the method for obtaining the period difference includes:
[0036] For any local signal segment, each other local signal segment of the local signal segment is sequentially used as an analysis signal segment;
[0037] For any signal value of the principal component of the local signal segment, obtain the number of distributions of the signal value in the principal component of the local signal segment as the target distribution number; obtain the number of distributions of the signal value in the principal component of the analysis signal segment as the analysis distribution number;
[0038] The difference between the target distribution quantity and the analysis distribution quantity is taken as the signal value distribution difference between the local signal segment and the analysis signal segment at the signal value; the accumulated value of the signal value distribution difference between the local signal segment and the analysis signal segment at all signal values is taken as the distribution deviation between the local signal segment and the analysis signal segment;
[0039] The average value of the distribution deviations between the local signal segment and all other local signal segments is taken as the period difference of the local signal segment.
[0040] Furthermore, the method for obtaining the interference indicator includes:
[0041] The product of the normalized value of the local interference confidence of each local signal segment and the interference degree of the state monitoring signal is used as the interference index of each local signal segment.
[0042] Furthermore, the state monitoring signal is enhanced based on the interference index of each local signal segment in the state monitoring signal to obtain the final monitoring signal, including:
[0043] The interference index of each local signal segment is used as the standard deviation in the convolution kernel of the Gaussian filter, and each local signal segment is Gaussian filtered to obtain a local enhanced signal; the local enhanced signals are re-spliced to obtain the final monitoring signal.
[0044] The present invention has the following beneficial effects:
[0045] The present invention is based on the correlation influence of various parameter monitoring signals of equipment operation and the state monitoring signal of the equipment, as well as the correlation between different types of parameter monitoring signals, to obtain the response of the state monitoring signal to the parameter monitoring signal, thereby reflecting the degree of influence of the equipment operation state on different operating environments, and then combining the fluctuation of the state monitoring signal and the parameter monitoring signal itself, improving the accuracy of the analysis of the interference of the state monitoring signal itself, and obtaining a more accurate interference degree. At the same time, considering the complex changes of various operating environments of the equipment, the interference influences in different local time periods may be different, considering the characteristics that better signal decomposition will be more beneficial to subsequent monitoring and analysis, the local signal segments divided by the state monitoring signal are divided by the degree of conformity of signal decomposition, and the data distribution deviation between different local signal segments, to obtain the local interference confidence of each local signal segment, reflecting the degree of participation of each local signal segment in subsequent interference removal enhancement. Finally, through the interference degree and the local interference confidence of the local signal segment, the degree of interference removal of each local signal segment is analyzed, and the interference index is obtained for subsequent segmented interference removal enhancement, so as to obtain a better final monitoring signal. The present invention analyzes the interference effects of various operating environments while considering the interference degree of the equipment status monitoring signal itself, and performs local adaptive enhancement through the different influences of local signal segments to obtain a higher quality monitoring signal and improve the accuracy of data analysis. BRIEF DESCRIPTION OF THE DRAWINGS
[0046] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0047] Figure 1 A structural diagram of an industrial park data integration and analysis system based on cloud computing provided by an embodiment of the present invention;
[0048] Figure 2 A schematic diagram of a status monitoring signal of a device provided by an embodiment of the present invention;
[0049] Figure 3 A flow chart of a method for obtaining the responsiveness of a state monitoring signal to each parameter monitoring signal provided by an embodiment of the present invention;
[0050] Figure 4 A schematic diagram of an ICA decomposition result of a state monitoring signal provided by an embodiment of the present invention;
[0051] Figure 5A flow chart of a method for obtaining local interference confidence of a local signal segment provided by an embodiment of the present invention;
[0052] Figure 6 A schematic diagram of a final monitoring signal provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0053] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the specific implementation method, structure, features and effects of a cloud computing-based industrial park data integration and analysis system proposed by the present invention are described in detail below in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.
[0054] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
[0055] The following is a detailed description of a specific solution of an industrial park data integration and analysis system based on cloud computing provided by the present invention in conjunction with the accompanying drawings.
[0056] See also Figure 1 , which shows a structural diagram of an industrial park data integration and analysis system based on cloud computing provided by an embodiment of the present invention, the system includes: a data acquisition module 601, an interference impact analysis module 602, a local interference analysis module 603 and an enhanced monitoring module 604.
[0057] The data acquisition module 601 is used to acquire the status monitoring signal of the equipment and the parameter monitoring signal of each operating environment during the preset time period of each equipment operation in the industrial park.
[0058] In today's industrial parks, in order to better monitor, manage and make decisions on data, the real-time and high concurrency performance of cloud computing technology is used to achieve online real-time data processing. Through data stream processing and instant analysis algorithms, the data changes of various operating equipment in the industrial park are monitored in real time, and timely abnormal detection warnings are provided. The operating status data of each device will not only be affected by the electromagnetic interference fluctuations during its own collection, but also by various operating environment changes. Therefore, data denoising and enhancement is an important data processing before accurate analysis.
[0059] Among them, various industrial data of the industrial park are collected through sensors. The sensors include various types to collect various status information. The sensors use network connections, such as the Internet of Things protocol, to transmit the collected industrial data to cloud servers or data centers for storage. According to data processing and analysis needs, the data is then transmitted to the industrial park data integration and analysis system for subsequent processing and analysis of the collected industrial data.
[0060] As an example, in the preset time period of each device in the industrial park, the sensors collect the time series data of the device's operating status and the time series data of different operating environments, and convert the collected time series data into signals to obtain status monitoring signals and various parameter monitoring signals. It should be noted that the conversion into signals enables the computer system to process and analyze efficiently, and facilitates database storage and network transmission, improving data analysis and data mining of device status and performance. Signal conversion is a technical means well known to those skilled in the art, such as conversion through analog-to-digital converters, etc., which will not be elaborated here. Please refer to Figure 2 , which shows a schematic diagram of a status monitoring signal of a device provided by an embodiment of the present invention.
[0061] In a specific implementation of the embodiment of the present invention, the preset time period can be set to 3 hours, and the specific time can be adjusted by the implementer. Among them, the operating status data can be selected from the operating voltage data of the equipment as an example, and other options include vibration data, load data or operating current of the equipment, etc. The types of operating environment may include: temperature, humidity or pressure, etc. For different equipment, the types of operating environment influences are also different. For example, in the transmission equipment, the number of conveyed objects on the conveyor belt is also a kind of operating environment data, which is obtained through video counting statistics. When the number of conveyed objects is large, it will also affect the operating status of the transmission equipment.
[0062] At this point, the acquisition of monitoring signal data is completed.
[0063] The interference impact analysis module 602 is used to obtain the responsiveness of the status monitoring signal to each parameter monitoring signal based on the degree of correlation between the status monitoring signal of the equipment and the parameter monitoring signal of each operating environment, as well as the correlation between each operating environment and other operating environments in the parameter monitoring signal; and to obtain the interference degree of the status monitoring signal based on the responsiveness of the status monitoring signal of the equipment to each parameter monitoring signal, as well as the stable difference in fluctuation between the status monitoring signal and the parameter monitoring signal of each operating environment.
[0064] The real-time impact of different operating environments on the status monitoring signal is different, and there are also correlations between some operating environments. For example, when the temperature of the environment is high, the resistance of the equipment will increase, causing voltage instability. At the same time, when the air pressure is low, the heat dissipation performance of the equipment will be affected due to the thin air, which will increase the temperature of the operating environment and bring unstable interference. Therefore, through the degree of correlation between the parameter monitoring signal and the status monitoring signal of the operating environment, as well as the possible correlation between different operating environments, it can reflect the responsiveness of the status monitoring signal to the parameter monitoring signal of each operating environment, that is, the degree of influence of the data changes of each operating environment on the status monitoring signal. Further based on the response situation combined with the instability of the status monitoring signal and the parameter monitoring signal itself, it is possible to comprehensively consider the impact of environmental fluctuations and characterize the degree of interference changes in the status monitoring signal due to various environmental interferences.
[0065] See also Figure 3 , which shows a flow chart of a method for obtaining the responsiveness of a state monitoring signal to each parameter monitoring signal provided by an embodiment of the present invention. Preferably, in some implementations of the embodiments of the present invention, the responsiveness obtaining method is as follows: Figure 3 As shown, the specific steps include:
[0066] S201: Obtain a change correlation degree of each parameter monitoring signal according to a data change correlation between a parameter monitoring signal of each operating environment and a status monitoring signal of a device.
[0067] The more consistent the data change of the parameter monitoring signal is with the change of the state monitoring signal, the more likely the state monitoring signal is to be affected by the parameter monitoring signal. Therefore, the influence of each parameter monitoring signal is reflected by the correlation of data changes. Each parameter monitoring signal is also the parameter monitoring signal of each operating environment.
[0068] Preferably, in some implementations of the embodiments of the present invention, the method for acquiring the change correlation degree includes:
[0069] The Pearson correlation coefficient between each parameter monitoring signal and the state monitoring signal is calculated and the absolute value is taken to obtain the change correlation of each parameter monitoring signal. The higher the absolute value of the Pearson correlation coefficient, the stronger the correlation between the data changes.
[0070] In other embodiments of the present invention, variance decomposition, such as partial least squares (PLS), may be used to analyze the variance contribution of different parameter monitoring signals to the state monitoring signal, and the coefficient of the parameter monitoring signal in the decomposition result is used as the change correlation. The larger the coefficient, the greater the contribution of the parameter monitoring signal in the component, the greater the impact on the state monitoring signal, and therefore the greater the change correlation.
[0071] In other embodiments of the present invention, regression analysis can also be used to establish a mathematical relationship between the state monitoring signal and each parameter monitoring signal. The larger the regression coefficient, the greater the correlation between the data change of the parameter monitoring signal and the state monitoring signal, and the regression coefficient is used as the change correlation.
[0072] It should be noted that the Pearson correlation coefficient, variance decomposition and regression analysis to establish mathematical relationships are all technical means well known to those skilled in the art and will not be elaborated here.
[0073] S202: Obtaining the influence representation degree of each parameter monitoring signal according to the change consistency between the parameter monitoring signal of each operating environment and all other parameter monitoring signals.
[0074] For each operating environment, when the change of this parameter monitoring signal is consistent with the changes of more parameter monitoring signals, the change characterization ability of this parameter monitoring signal will become weaker. The impact of the monitoring signal can be reflected by the changes in other operating environments. If this operating environment also participates in the analysis of the impact of interference on the status monitoring signal, it will cause repeated analysis of the interference, affect the subsequent enhancement of the status monitoring signal, and increase the error. Therefore, it is necessary to characterize the impact characterization degree of each parameter monitoring signal in combination with the consistency of the signal change.
[0075] Preferably, in some implementations of the embodiments of the present invention, the method for obtaining the influence representation degree includes:
[0076] First, the parameter monitoring signal of each operating environment is standardized to obtain a standard curve for each parameter monitoring signal. The standard curve analysis after unified dimensioning facilitates the analysis of the approximate fluctuation of the numerical values of the parameter monitoring signal. For any parameter monitoring signal, the mean square error between the standard curve of the parameter monitoring signal and the standard curve of each other parameter monitoring signal is used as the approximation of the numerical distribution of the parameter monitoring signal and the other monitoring signals. The mean square error reflects the approximation of the fluctuation values of the two signal data. In other embodiments of the present invention, the mean absolute error between the standard curves can also be used as the approximation of the numerical distribution, which will not be elaborated here.
[0077] Secondly, the standard curve of the parameter monitoring signal and the DTW value of each other parameter monitoring signal are used as the approximation of the numerical change between the parameter monitoring signal and each other parameter monitoring signal. The DTW value obtained by the dynamic time warping (DTW) algorithm reflects the similarity of the two signal data in the overall change. In other embodiments of the present invention, the mutual information between the parameter monitoring signal and each other parameter monitoring signal can also be obtained as the approximation of the numerical change, and the correlation and dependence between the parameter monitoring signal and each other parameter monitoring signal can be evaluated by the mutual information, which will not be elaborated here.
[0078] Finally, the influence characterization of the parameter monitoring signal is obtained by combining the numerical distribution approximation and numerical change approximation of the parameter monitoring signal with all other parameter monitoring signals, and both the numerical distribution approximation and the numerical change approximation are negatively correlated with the influence characterization. As an example, the product of the numerical distribution approximation and the numerical change approximation of the parameter monitoring signal with each other parameter monitoring signal is negatively correlated to obtain the influence characterization of the parameter monitoring signal with each other parameter monitoring signal. The negative correlation mapping can be implemented in the form of an inverse proportional value or a negative exponential power, which is not limited here.
[0079] S203: Determine the responsiveness of the state monitoring signal to each parameter monitoring signal through the change correlation degree of each parameter monitoring signal and the influence characterization degree of each parameter monitoring signal.
[0080] By adjusting the degree to which each parameter monitoring signal represents the state monitoring signal when it changes, the change correlation of each parameter monitoring signal is reflected to reflect the response degree of the state monitoring signal to the change of each parameter monitoring signal.
[0081] In one embodiment of the present invention, the influence representation degree of each parameter monitoring signal is normalized and multiplied by the change correlation to obtain the responsiveness of the state monitoring signal to each parameter monitoring signal. It should be noted that normalization is a technical means well known to those skilled in the art, and the normalization function can be selected as linear normalization or standard normalization, etc. The specific normalization method is not limited here.
[0082] At this point, the acquisition of responsiveness is completed.
[0083] In the analysis of interference conditions, the difference in signal values between the monitored signal and the expected signal is used for analysis. For the state monitoring signal, it is itself affected by the electromagnetic field, mechanical vibration or environment, resulting in complex interference fluctuations. These fluctuations have a great impact on normal state monitoring and can easily lead to misjudgment of abnormalities. Therefore, the degree of interference is analyzed through the operating parameters and responsiveness, as well as the fluctuations of the state monitoring itself. For parameters with higher responses, the more abnormal the fluctuations of the monitoring signal are, the more likely it is that the abnormal difference in the state monitoring signal is caused by the operating environment, and interference analysis is performed.
[0084] Preferably, in some implementations of the embodiments of the present invention, obtaining the interference degree of the status monitoring signal includes:
[0085] First, according to the difference between the parameter monitoring signal of each operating environment and the corresponding preset parameter expected signal, the abnormality of each parameter monitoring signal is determined. As an example, the difference in signal value between each parameter monitoring signal and the corresponding preset parameter expected signal at each moment is taken as the abnormal difference of each parameter monitoring signal at each moment, and the sum of all abnormal differences is taken as the abnormality of each parameter monitoring signal. The preset parameter expected signal is the ideal mode of the signal when there is no abnormal change. According to the operation setting of the specific implementation scenario, no restriction is made here. The accumulation of differences reflects the overall difference between the signal and the normal state. The greater the abnormality, the greater the change in the corresponding type of operating environment data.
[0086] Furthermore, the possibility of interference of the state monitoring signal is obtained by combining the responsiveness of the state monitoring signal to each parameter monitoring signal and the abnormality of each parameter monitoring signal. As an example, for the parameter monitoring signal of any operating environment, the responsiveness of the state monitoring signal to the parameter monitoring signal is multiplied by the abnormality of the parameter monitoring signal to obtain the influence interference of the parameter monitoring signal; the average value of the impression interference of the parameter monitoring signals of all operating environments is normalized to obtain the possibility of interference of the state monitoring signal. For each parameter monitoring signal, when the abnormality is greater and the responsiveness is higher, it means that the abnormality of the parameter monitoring signal brings greater fluctuations to the state monitoring signal, and the state monitoring signal is more disturbed by the parameter monitoring signal, so the influence interference is higher. Then, the influence interference of all types of parameter monitoring signals is combined to comprehensively analyze the possibility of interference of the state monitoring signal.
[0087] Furthermore, based on the difference between the state monitoring signal and the preset state expected signal, the difference degree of the state monitoring signal is determined. As an example, the difference in signal value between the state monitoring signal and the corresponding preset state expected signal at each moment is taken as the deviation degree of the state monitoring signal at each moment, and the average value of all deviations is taken as the difference degree of the state monitoring signal, reflecting the degree of deviation of the state monitoring signal from normal monitoring. This deviation degree may include abnormal signal parts and disturbed signal parts.
[0088] Finally, the interference degree of the state monitoring signal is obtained according to the difference degree and the interference possibility degree of the state monitoring signal. As an example, the product of the difference degree of the state monitoring signal and the interference possibility degree is used as the interference degree of the state monitoring signal, so as to realize the analysis and judgment of the interference situation in the difference degree.
[0089] The local interference analysis module 603 is used to divide the status monitoring signal of the equipment into at least two local signal segments; obtain the signal segment components of each local signal segment after signal decomposition; and obtain the local interference confidence of each local signal segment based on the periodic deviation of the data distribution in the signal segment components between each local signal segment and other local signal segments, and the degree to which each signal segment component in each local signal segment complies with the decomposition rules.
[0090] In the monitoring time period, the interference degree of the status monitoring signal in different local segments is different, and there are different differences in the interference conditions. Therefore, the status monitoring signal is divided and analyzed to divide the status monitoring signal of the equipment into at least two local signal segments.
[0091] In a specific implementation of the embodiment of the present invention, the state monitoring signal is evenly divided into a preset number of local signal segments, and the preset number is set to 20, which can be adjusted by the implementer according to the specific implementation situation.
[0092] For each local signal segment, it is difficult to analyze the situation of the local signal segment due to the complexity of the superimposed state monitoring signal. Therefore, the complex situation of the local signal segment is simplified through signal decomposition, and the possibility of interference in the local signal segment can be analyzed through the decomposition situation.
[0093] In a specific implementation of the embodiment of the present invention, the signal is decomposed by independent component analysis (ICA) to obtain the signal segment components of each local signal segment after the signal is decomposed. It should be noted that the ICA separation method is a technical means well known to those skilled in the art, and Fourier transform, wavelet transform or empirical mode decomposition can also be used, which will not be described in detail here. Please refer to Figure 4 , which shows a schematic diagram of an ICA decomposition result of a state monitoring signal provided by an embodiment of the present invention.
[0094] When the signal is disturbed, these interferences may appear as additional or statistically related components to the original signal. Therefore, during ICA decomposition, these interference components may destroy the assumption of ICA decomposition, that is, the original signal is statistically independent, so interference may cause the result of ICA decomposition to be inaccurate, making the decomposed components not meet the rules of independent decomposition. In other embodiments of the present invention, when wavelet decomposition is adopted, the interference part will also destroy the independence of components of different scales, resulting in high component correlation and poor decomposition effect.
[0095] At the same time, since the data distribution of the signal in the normal operating state has consistent periodicity, the higher the possibility of data distribution with periodic consistency in the decomposed components, the higher the stability of different components in the superimposed signal, and the lower the possibility of interference in this local signal segment.
[0096] Therefore, for each signal segment component, the interference degree of each local signal segment is analyzed from two aspects: period distribution and decomposition rule compliance. Figure 5 , which shows a flow chart of a method for obtaining a local interference confidence of a local signal segment provided by an embodiment of the present invention. Preferably, in some implementations of the embodiments of the present invention, the method for obtaining the local interference confidence is as follows: Figure 5 As shown, the specific steps include:
[0097] S301: determining a principal component among all signal segment components of each local signal segment; and determining a period difference degree of each local signal segment according to a distribution degree of a signal value of the principal component in each local signal segment among principal components of other local signal segments.
[0098] The principal components of the signal in the local signal segment are periodically analyzed through the principal component. When the signal value distribution of the principal component in a single local signal segment is greatly different from that of the principal components in other local signal segments, it indicates that there is obvious interference in this local signal segment.
[0099] In a specific implementation of an embodiment of the present invention, a principal component analysis (PCA) method is used to determine the principal component of each local signal segment. The PCA method is a technical means well known to those skilled in the art and will not be described in detail here.
[0100] Preferably, in some implementations of the embodiments of the present invention, the method for obtaining the period difference includes:
[0101] For any local signal segment, each other local signal segment of the local signal segment is used as the analysis signal segment in turn; for any signal value of the main component of the local signal segment, the number of distributions of the signal value in the main component of the local signal segment is obtained as the target distribution number; the number of distributions of the signal value in the main component of the analysis signal segment is obtained as the analysis distribution number. By obtaining the distribution number of a single signal value, the distribution of the signal value in the main component can be reflected, and then the difference analysis can be performed to reflect the consistency of the distribution. The more periodic the distribution of the main components of the two local signal segments is, the more consistent the distribution number of the signal values will be.
[0102] Furthermore, the difference between the target distribution number and the analysis distribution number is taken as the signal value distribution difference between the local signal segment and the analysis signal segment at the signal value; the cumulative value of the signal value distribution difference between the local signal segment and the analysis signal segment at all signal values is taken as the distribution deviation between the local signal segment and the analysis signal segment. The larger the difference, that is, the larger the difference in signal value distribution, the worse the distribution regularity of the principal components of the local signal segment and the analysis signal segment at a single signal value, and then combined with the difference of all signal values, the distribution deviation is obtained. The larger the distribution deviation, the larger the deviation in periodic distribution of the principal components of the local signal segment and the analysis signal segment.
[0103] Furthermore, the average value of the distribution deviation between the local signal segment and all other local signal segments is used as the period difference of the local signal segment. Based on the overall deviation of the principal component between the local signal segment and other local signal segments, the larger the period difference, the more inconsistent the distribution of the principal component data between the local signal segment and other local signal segments is, and the greater the degree of interference in the local signal segment.
[0104] In other embodiments of the present invention, the local signal segment and the analysis signal segment may be subjected to Fourier transform to obtain a spectrum, and the difference in amplitude corresponding to the same frequency between the spectrum of the local signal segment and the analysis signal segment is calculated, and the sum is taken to obtain the period difference between the local signal segment and the analysis signal segment. The difference in periodic patterns between signals is analyzed in the form of a spectrum diagram, and the Fourier transform of spectrum conversion is a technical means well known to those skilled in the art, and will not be described in detail here.
[0105] S302: Determine the decomposition regularity of each local signal segment based on the Gaussianity of each signal segment component in each local signal segment.
[0106] In the decomposition rules of ICA decomposition, in order to separate components with strong independence, components with strong non-Gaussianity are usually found to better capture independent sources in the original signal. When a component is close to Gaussian analysis, it reflects that this component may contain more similar information to other components and is not unique. It is very likely caused by interference. Therefore, the decomposition regularity is obtained through Gaussianity.
[0107] As an example, for a signal segment component in any local signal segment, the kurtosis of the signal segment component is obtained; the difference between the kurtosis of the signal segment and the kurtosis of the standard normal distribution is negatively correlated to obtain the Gaussianity of the signal segment; the average value of the Gaussianity of all signal segment components in the local signal segment is taken as the decomposition regularity of the local signal segment.
[0108] The kurtosis can reflect the Gaussian degree of the signal. When the kurtosis of the overall signal segment component in the local signal segment is closer to the kurtosis of the standard normal distribution, that is, when the decomposition regularity is greater, it means that the Gaussian nature of the component is stronger, and the decomposition effect of the local signal segment is worse, and the interference may be higher. It should be noted that the kurtosis is a technical means well known to those skilled in the art, and the kurtosis of the standard normal distribution is 3, which will not be described in detail here. In other embodiments of the present invention, the negative entropy can also be used to reflect the Gaussian degree of the signal, which will not be described in detail here.
[0109] S303: Obtaining the local interference confidence of each local signal segment according to the period difference and decomposition regularity of each local signal segment; both the period difference and the decomposition regularity are positively correlated with the local interference confidence.
[0110] The interference in the local signal segment is analyzed by combining the periodic difference and decomposition regularity. The larger the periodic difference, the larger the data distribution difference in the component, the weaker the signal periodic feature, and the more likely the local signal segment is to have greater interference. The larger the decomposition regularity, the poorer the signal decomposition effect of the local signal segment, and the more likely the local signal segment is to have greater interference. Therefore, the periodic difference and decomposition regularity are positively correlated with the local interference confidence.
[0111] As an example, the product of the period difference and decomposition regularity of each local signal segment is used as the local interference confidence of each local signal segment. The local interference confidence reflects the degree of interference in the local signal segment. When the local interference confidence is larger, it means that the local signal segment may be affected by greater interference and needs stronger enhancement in the future.
[0112] At this point, the analysis of each local signal segment is completed.
[0113] The enhanced monitoring module 604 is used to determine the interference index of each local signal segment through the interference degree of the state monitoring signal and the local interference confidence of each local signal segment; enhance the state monitoring signal based on the interference index of each local signal segment in the state monitoring signal to obtain the final monitoring signal.
[0114] In combination with the local interference situation and the overall interference degree of the signal, the interference situation of each local signal segment is obtained. Preferably, in some implementations of the embodiments of the present invention, the method for obtaining the interference indicator includes:
[0115] The product of the normalized value of the local interference confidence of each local signal segment and the interference degree of the state monitoring signal is used as the interference index of each local signal segment. The interference index reflects the degree of interference of each local signal segment.
[0116] The state monitoring signal can be further enhanced based on the interference index of each local signal segment in the state monitoring signal to obtain the final monitoring signal. As an example, the interference index of each local signal segment is used as the standard deviation in the convolution kernel of the Gaussian filter, and each local signal segment is Gaussian filtered to obtain a local enhanced signal, and the local enhanced signals are re-joined to obtain the final monitoring signal. Figure 6 , which shows a schematic diagram of a final monitoring signal provided by an embodiment of the present invention.
[0117] Ultimately, the monitoring signal achieves local adaptive enhancement of the status monitoring signal, improves the quality of the monitoring signal, and then evaluates and warns the equipment status through the industrial park data integration and analysis system, so as to better detect and maintain the equipment in the industrial park, and provide industrial park managers with more comprehensive and accurate data support for better monitoring, management and decision-making.
[0118] In summary, the present invention is based on the correlation influence of various parameter monitoring signals of equipment operation and the state monitoring signal of the equipment, as well as the correlation between different types of parameter monitoring signals, to obtain the response of the state monitoring signal to the parameter monitoring signal, thereby reflecting the degree of influence of the equipment operation state on different operating environments, and then combining the fluctuation of the state monitoring signal and the parameter monitoring signal itself, improving the accuracy of the analysis of the interference of the state monitoring signal itself, and obtaining a more accurate interference degree. At the same time, considering the complex changes of various operating environments of the equipment, the interference influences in different local time periods may be different, considering that the better signal decomposition will be more beneficial to the subsequent monitoring and analysis, the local signal segments divided by the state monitoring signal are obtained by the degree of conformity of the signal decomposition, and the data distribution deviation between different local signal segments, and the local interference confidence of each local signal segment is obtained, reflecting the degree of participation of each local signal segment in the subsequent interference removal enhancement. Finally, through the interference degree and the local interference confidence of the local signal segment, the degree of interference removal of each local signal segment is analyzed, and the interference index is obtained for subsequent segmented interference removal enhancement, so as to obtain a better final monitoring signal. The present invention analyzes the interference effects of various operating environments while considering the interference degree of the equipment status monitoring signal itself, and performs local enhancement through the different influences of local signal segments to obtain a higher quality monitoring signal and improve the accuracy of data analysis.
[0119] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
[0120] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.
Claims
1. An industrial park data integration and analysis system based on cloud computing, characterized in that: The system comprises: The data acquisition module is used to obtain the status monitoring signal of the equipment and the parameter monitoring signal of each operating environment during the preset time period of each equipment operation in the industrial park; The interference impact analysis module is used to obtain the responsiveness of the state monitoring signal to each parameter monitoring signal according to the degree of correlation between the state monitoring signal of the equipment and the parameter monitoring signal of each operating environment, and the correlation between each operating environment and other operating environments in the parameter monitoring signal; obtain the interference degree of the state monitoring signal according to the responsiveness of the state monitoring signal of the equipment to each parameter monitoring signal, and the stable difference in fluctuation between the state monitoring signal and the parameter monitoring signal of each operating environment; A local interference analysis module is used to divide the state monitoring signal of the equipment into at least two local signal segments; obtain the signal segment components of each local signal segment after signal decomposition; obtain the local interference confidence of each local signal segment according to the periodic deviation of the data distribution in the signal segment components between each local signal segment and other local signal segments, and the degree to which each signal segment component in each local signal segment complies with the decomposition rule; The enhanced monitoring module is used to determine the interference index of each local signal segment through the interference degree of the state monitoring signal and the local interference confidence of each local signal segment; enhance the state monitoring signal based on the interference index of each local signal segment in the state monitoring signal to obtain the final monitoring signal.
2. According to the cloud computing-based industrial park data integration and analysis system of claim 1, it is characterized in that: The method for obtaining the responsiveness of the state monitoring signal to each parameter monitoring signal includes: According to the data change correlation between the parameter monitoring signal of each operating environment and the status monitoring signal of the equipment, the change correlation degree of each parameter monitoring signal is obtained; According to the change consistency between the parameter monitoring signal of each operating environment and all other parameter monitoring signals, the influence characterization degree of each parameter monitoring signal is obtained; The responsiveness of the state monitoring signal to each parameter monitoring signal is determined by the change correlation degree of each parameter monitoring signal and the influence characterization degree of each parameter monitoring signal.
3. According to claim 2, a cloud computing-based industrial park data integration and analysis system is characterized in that: The obtaining of the influence characterization degree of each parameter monitoring signal according to the change consistency of the parameter monitoring signal of each operating environment and all other parameter monitoring signals includes: Standardize the parameter monitoring signal of each operating environment to obtain a standard curve for each parameter monitoring signal; For any parameter monitoring signal, the mean square error between the standard curve of the parameter monitoring signal and the standard curve of each other parameter monitoring signal is taken as the numerical distribution approximation degree of the parameter monitoring signal and the other parameter monitoring signals; The standard curve of the parameter monitoring signal and the DTW value of each other parameter monitoring signal are used as the numerical change approximation of the parameter monitoring signal and each other parameter monitoring signal; The influence characterization of this parameter monitoring signal is obtained by combining the numerical distribution approximation and numerical change approximation of this parameter monitoring signal with all other parameter monitoring signals. Both the numerical distribution approximation and the numerical change approximation are negatively correlated with the influence characterization.
4. According to the cloud computing-based industrial park data integration and analysis system of claim 1, it is characterized in that: The method for obtaining the interference degree includes: Determine the abnormality of each parameter monitoring signal according to the difference between the parameter monitoring signal of each operating environment and the corresponding preset parameter expected signal; Combining the responsiveness of the state monitoring signal to each parameter monitoring signal and the abnormality of each parameter monitoring signal, the possibility of interference of the state monitoring signal is obtained; Determining the difference degree of the state monitoring signal according to the difference between the state monitoring signal and the preset state expected signal; The interference degree of the state monitoring signal is obtained according to the difference degree and interference possibility degree of the state monitoring signal.
5. According to claim 4, a cloud computing-based industrial park data integration and analysis system is characterized in that: The method for obtaining the abnormality degree includes: The difference in signal value between each parameter monitoring signal at each moment and the corresponding preset parameter expected signal is taken as the abnormal difference of each parameter monitoring signal at each moment; The sum of all abnormal differences is taken as the abnormality of each parameter monitoring signal.
6. The cloud computing-based industrial park data integration and analysis system according to claim 4 is characterized in that: The method for obtaining the possibility of interference includes: For any parameter monitoring signal of an operating environment, the responsiveness of the state monitoring signal to the parameter monitoring signal is multiplied by the abnormality of the parameter monitoring signal to obtain the influence interference degree of the parameter monitoring signal; The average values of the impression interference of the parameter monitoring signals of all operating environments are normalized to obtain the interference possibility of the state monitoring signal.
7. The cloud computing-based industrial park data integration and analysis system according to claim 1 is characterized in that: The method for obtaining the local interference confidence comprises: determining a principal component among all signal segment components of each local signal segment; Determine the period difference of each local signal segment according to the distribution degree of the signal value of the principal component in each local signal segment in the principal components of other local signal segments; Determining the decomposition regularity of each local signal segment based on the Gaussianity of each signal segment component in each local signal segment; According to the period difference and decomposition regularity of each local signal segment, the local interference confidence of each local signal segment is obtained; the period difference and decomposition regularity are both positively correlated with the local interference confidence.
8. The cloud computing-based industrial park data integration and analysis system according to claim 7 is characterized in that: The method for obtaining the period difference comprises: For any local signal segment, each other local signal segment of the local signal segment is sequentially used as an analysis signal segment; For any signal value of the principal component of the local signal segment, obtain the number of distributions of the signal value in the principal component of the local signal segment as the target distribution number; obtain the number of distributions of the signal value in the principal component of the analysis signal segment as the analysis distribution number; The difference between the target distribution quantity and the analysis distribution quantity is taken as the signal value distribution difference between the local signal segment and the analysis signal segment at the signal value; the accumulated value of the signal value distribution difference between the local signal segment and the analysis signal segment at all signal values is taken as the distribution deviation between the local signal segment and the analysis signal segment; The average value of the distribution deviations between the local signal segment and all other local signal segments is taken as the period difference of the local signal segment.
9. The cloud computing-based industrial park data integration and analysis system according to claim 1 is characterized in that: The method for obtaining the interference index includes: The product of the normalized value of the local interference confidence of each local signal segment and the interference degree of the state monitoring signal is used as the interference index of each local signal segment.
10. The cloud computing-based industrial park data integration and analysis system according to claim 1, characterized in that: The step of enhancing the state monitoring signal based on the interference index of each local signal segment in the state monitoring signal to obtain a final monitoring signal includes: The interference index of each local signal segment is used as the standard deviation in the convolution kernel of the Gaussian filter, and each local signal segment is Gaussian filtered to obtain a local enhanced signal; the local enhanced signals are re-spliced to obtain the final monitoring signal.
