A method and system for intelligent diagnosis of reactor faults for acidic reactions

Abstract

The present invention relates to the technical field of vibration signal fault diagnosis, and proposes an intelligent fault diagnosis method and system for a reactor for acidic reactions, including: collecting vibration signals of the bearings of the reactor; decomposing the vibration signals into component signals, determining the fluctuation degree and independence of the component signals, and obtaining fault characteristic values ​​of the component signals of the vibration signals; determining the local energy of each component signal at any acquisition time according to the amplitude of the vibration signal of the same component signal at different acquisition times, and determining the signal fault degree of the component signal of the vibration signal according to the fault characteristic value of the component signal of the vibration signal and the difference in the local energy of the component signal of the vibration signal at two adjacent acquisition times; determining abnormal signals according to the signal fault degrees of all component signals of the vibration signal, and obtaining reactor fault diagnosis results. The present invention solves the problem that composite faults may cause mutual interference between vibration signals, resulting in inaccurate fault diagnosis results of the reactor bearings.

Description

Technical Field

[0001] The present invention relates to the technical field of vibration signal fault diagnosis, and in particular to an intelligent fault diagnosis method and system for a reactor for acidic reactions. Background Art

[0002] The reactor is a chemical reaction container, generally composed of a reactor body, a rotating device, a stirring device, a heating device, a cooling device and a sealing device. The bearing is a key component of the stirring device of the reactor, which directly affects the reliability and stability of the operation of the reactor. During the operation of the reactor, the bearing of the reactor is subjected to high temperature, high pressure and acidic environment for a long time, and is prone to failure. Therefore, it is necessary to diagnose the fault of the reactor bearing in time to avoid further damage to the reactor.

[0003] The failure of the reactor's bearing is often not a single failure, but a composite failure that causes mutual interference between the vibration signals generated by the reactor's bearing. Therefore, using only traditional signal analysis algorithms to detect bearing failures is often affected by the mutual interference between vibration signals caused by composite failures, resulting in inaccurate intelligent diagnosis results for the reactor's bearing failures, making it difficult for the intelligent diagnosis results to reflect the true health status of the reactor's bearings. Summary of the invention

[0004] The present invention provides an intelligent fault diagnosis method and system for an acidic reaction reactor to solve the problem that composite faults may cause mutual interference between vibration signals, resulting in inaccurate fault diagnosis results for the reactor bearing. The technical solutions adopted are as follows:

[0005] In a first aspect, an embodiment of the present invention provides a method for intelligent diagnosis of reactor faults for acidic reactions, the method comprising the following steps:

[0006] Collect vibration signals of the reactor's bearings at different collection times;

[0007] Decomposing the vibration signal into component signals of a preset number of frequency bands, determining the degree of fluctuation of the component signals of the vibration signal according to the difference between the component signals of the vibration signal and the fluctuation of the amplitude contained in the vibration signal, determining the independence of each component signal of the vibration signal according to the correlation between different component signals of the vibration signal, and obtaining the fault characteristic value of the component signal of the vibration signal according to the degree of fluctuation and independence of the component signals of the vibration signal;

[0008] According to the amplitude values ​​of the vibration signal of the same component signal at different acquisition moments, the local energy of each component signal at any acquisition moment is determined, and according to the fault characteristic value of the component signal of the vibration signal and the difference in the local energy of the component signal of the vibration signal at two adjacent acquisition moments, the signal fault degree of the component signal of the vibration signal is determined;

[0009] Determine whether there is an abnormal signal based on the signal fault degree of all component signals of the vibration signal, and obtain the reactor fault diagnosis result.

[0010] Furthermore, the specific method for obtaining the fluctuation degree of the component signal of the vibration signal is as follows:

[0011] The information entropy of the amplitude contained in the component signal of the vibration signal is recorded as the first information entropy of the component signal, the information entropy of the amplitude contained in the vibration signal is recorded as the second information entropy of all component signals of the vibration signal, and the ratio of the first information entropy to the second information entropy of the component signal is recorded as the fluctuation degree of the component signal.

[0012] Further, the independence of each component signal of the vibration signal is determined according to the correlation between different component signals of the vibration signal, and the specific method includes:

[0013] Calculate the correlation between different component signals of the vibration signal, and record the normalized value of the correlation between the different component signals as the normalized correlation between the different component signals;

[0014] The independence of the component signals is obtained based on the normalized correlation between all different component signals.

[0015] Furthermore, the independence of the component signals is obtained according to the normalized correlation between all the different component signals, and the specific method includes:

[0016] Determining the difference between the different component signals according to the normalized correlation between the different component signals, wherein the difference between the different component signals is negatively correlated with the normalized correlation between the different component signals;

[0017] The sum of the differences between any component signal and all other component signals is recorded as the independence of the component signal.

[0018] Furthermore, the specific method for obtaining the fault characteristic value of the component signal of the vibration signal is as follows:

[0019] The product of the fluctuation degree and independence of the component signal of the vibration signal is recorded as the fault characteristic value of the component signal of the vibration signal.

[0020] Furthermore, the method of determining the local energy of each component signal at any acquisition time according to the amplitude values ​​of the vibration signal of the same component signal at different acquisition times includes the following specific methods:

[0021] Record any moment of vibration signal collection as the target moment;

[0022] For the same component signal, the sum of the squares of the amplitudes of the vibration signals at the time when the third preset constant number of vibration signals centered at the target time is collected is recorded as the local energy of the same component signal at the target time.

[0023] Further, the signal fault degree of the component signal of the vibration signal includes the following specific methods:

[0024] The absolute value of the difference between the local energy of the component signal of the vibration signal at the target time and the next adjacent time of the target time is recorded as the local energy difference of the component signal of the vibration signal at the target time;

[0025] The signal fault degree of the component signal is obtained according to the fault characteristic value of the component signal and the local energy difference of the component signal at all times.

[0026] Further, the signal fault degree of the component signal is obtained according to the fault characteristic value of the component signal and the local energy difference of the component signal at all times, including the specific method of:

[0027] The cumulative sum of the local energy differences of the component signals of the vibration signal at all times is recorded as the energy difference change of the component signals of the vibration signal;

[0028] The product of the energy difference change of the component signal of the vibration signal and the fault characteristic value is recorded as the signal fault degree of the component signal of the vibration signal.

[0029] Further, the method of determining whether there is an abnormal signal according to the signal fault degree of all component signals of the vibration signal and obtaining the reactor fault diagnosis result includes the following specific methods:

[0030] Perform anomaly detection on the signal fault degrees of all component signals of the vibration signal, obtain anomaly scores of the signal fault degrees of all component signals, and mark the component signals corresponding to the signal fault degrees whose anomaly scores are greater than or equal to an anomaly threshold as abnormal signals;

[0031] When an abnormal signal is identified, it is determined that the reactor is faulty; when no abnormal signal is identified, it is determined that the reactor is operating normally.

[0032] In a second aspect, an embodiment of the present invention further provides an intelligent fault diagnosis system for reactors for acidic reactions, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor implements the steps of any one of the above methods when executing the computer program.

[0033] The beneficial effects of the present invention are:

[0034] The present application evaluates the difference between the component signals of the vibration signal and the amplitude fluctuation contained in the vibration signal according to the characteristics that the amplitude of the vibration signal of the bearing with a composite fault will increase abnormally, and some component signals of specific frequencies that do not exist in normal working conditions or have very small amplitudes will appear, and obtain the fluctuation degree of the component signals. When the bearing has a composite fault, the amplitude of the vibration signal will increase abnormally, and some component signals of specific frequencies that do not exist in normal working conditions or have correspondingly very small amplitudes will appear, and the fluctuation degree of the component signals of the vibration signal is large. Then, combined with the characteristics that the composite fault of the bearing will cause multiple different fault signals to be superimposed on each other in the time domain and the frequency domain, resulting in an increase in the correlation between different component signals of the vibration signal, the fault characteristic value of the component signal of the vibration signal is obtained, and the fault characteristic value of the component signal is a significance evaluation of the composite fault characteristics presented by the component signal; further, combined with the characteristics that the energy of the vibration signal with a composite fault will increase significantly, the signal fault degree of the component signal of the vibration signal is determined, and whether there is an abnormal signal is determined according to the signal fault degree of all component signals of the vibration signal, and the fault diagnosis result of the reactor is obtained, so as to solve the problem that the composite fault will cause mutual interference between vibration signals, resulting in inaccurate fault diagnosis results of the reactor bearing. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be 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 labor.

[0036] Figure 1 A schematic diagram of a process flow of an intelligent diagnosis method for reactor faults in acidic reactions provided by an embodiment of the present invention;

[0037] Figure 2 A flowchart for obtaining fault characteristic values ​​provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0038] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0039] See also Figure 1 , which shows a flow chart of a method for intelligent diagnosis of reactor faults for acidic reactions provided by an embodiment of the present invention, the method comprising the following steps:

[0040] Step S001, collecting vibration signals of the bearings of the reactor at different collection times.

[0041] The vibration sensor is fixed on the bearing seat and used to collect the vibration signal generated by the bearing during the operation of the reactor.

[0042] This embodiment takes into account the Nyquist sampling theorem and the needs of practical applications, and sets the sampling frequency of the vibration signal to 48 kHz. As other implementation methods, the implementer can determine the sampling frequency of the vibration signal according to actual conditions, and this application does not impose any special restrictions.

[0043] The collected vibration signal is denoised using a median filter to reduce the impact of environmental noise on the collected vibration signal. The use of a median filter to denoise the vibration signal is a well-known technique and will not be described in detail.

[0044] At this point, the vibration signal is obtained.

[0045] Step S002, decomposing the vibration signal into component signals of a preset number of frequency bands, determining the fluctuation degree of the component signals of the vibration signal according to the difference between the component signals of the vibration signal and the fluctuation of the amplitude contained in the vibration signal, determining the independence of each component signal of the vibration signal according to the correlation between different component signals of the vibration signal, and obtaining the fault characteristic value of the component signal of the vibration signal according to the fluctuation degree and independence of the component signals of the vibration signal.

[0046] The bearings of the reactor are subjected to high temperature, high pressure and acidic environment for a long time, and are prone to compound faults. The faults of the reactor include inner ring fault, outer ring fault, rolling element fault and cage fault. The vibration signal of the bearing itself is a non-stationary nonlinear signal. Under the influence of compound faults, the vibration signals interfere with each other, and different vibration signals may overlap in the same frequency band. It is impossible to fully and accurately identify compound faults and obtain accurate intelligent diagnosis results of the reactor bearing faults by relying solely on frequency domain or time domain analysis.

[0047] Compared with the normal working state of the reactor bearing, when the bearing has a composite fault, the amplitude of the vibration signal will increase abnormally, and some component signals of specific frequencies that do not exist or have a very small corresponding amplitude will appear in the normal working state. For example, when the inner ring, outer ring or rolling element of the bearing is damaged, impact vibration will be generated, and the frequency components of these impact vibrations will be manifested in different frequency bands of the vibration signal. These impact vibrations do not exist or have a very small corresponding amplitude in the normal working state of the reactor bearing.

[0048] Using the empirical wavelet transform algorithm, the vibration signal is decomposed into component signals of a first preset constant frequency band. In this embodiment, the first preset constant is recorded as , the value of the first preset constant is 7, and the vibration signal is decomposed using the empirical wavelet transform algorithm, which is a well-known technology and will not be described in detail. The component signals are denoted as ,in, .

[0049] The degree of fluctuation of the component signal of the vibration signal is determined according to the difference between the component signal of the vibration signal and the fluctuation of the amplitude contained in the vibration signal.

[0050] Preferably, as an embodiment of the present application, the information entropy of the amplitude contained in the component signal of the vibration signal is recorded as the first information entropy of the component signal, the information entropy of the amplitude contained in the vibration signal is recorded as the second information entropy of all component signals of the vibration signal, and the ratio of the first information entropy to the second information entropy of the component signal is recorded as the fluctuation degree of the component signal.

[0051] Specifically, the calculation formula for the fluctuation degree of the component signal is:

[0052]

[0053] In the formula, The vibration signal The volatility of the component signals; The vibration signal component signals; The vibration signal The first information entropy of the component signals; The vibration signal The second information entropy of the component signals.

[0054] The fluctuation degree of the component signal of the vibration signal is an evaluation of the fluctuation of the component signal of the vibration signal relative to the amplitude contained in the vibration signal. When a bearing has a composite fault, the amplitude of the vibration signal will increase abnormally, and some component signals of specific frequencies that do not exist or have a very small corresponding amplitude will appear under normal working conditions. Therefore, the first information entropy of the component signal is larger than that when the bearing is working normally, and the fluctuation degree of the component signal of the vibration signal is larger.

[0055] It should be noted that the present embodiment selects information entropy to evaluate the fluctuation of the amplitude contained in the component signal of the vibration signal. On the basis of achieving the purpose of evaluating the fluctuation of the amplitude, in the actual application process, the implementer may use other existing methods such as approximate entropy, variance, etc. to evaluate the fluctuation of the amplitude, and the present application does not impose any special restrictions.

[0056] When a composite fault occurs in a bearing, the various fault signals contained in the composite fault are superimposed on each other in the time domain and frequency domain, which increases the correlation between the different component signals of the vibration signal and the complexity of the different component signals of the vibration signal. This results in inaccurate intelligent diagnosis results of the reactor bearing fault obtained by directly using traditional signal analysis algorithms to detect the bearing fault.

[0057] Calculate the correlation between different component signals of the vibration signal. When the correlation between a component signal and other component signals is greater, the component signal is more likely to be a component signal formed by the interaction of multiple faults. When the correlation between different component signals is smaller, the similarity between these different component signals is smaller, and they are more likely to correspond to different fault characteristics or normal component signals.

[0058] Preferably, this embodiment selects cosine similarity as the correlation measure between different component signals of the vibration signal. As other implementation methods, on the basis of achieving the purpose of measuring the correlation between two component signals, the implementer can adopt other methods in the prior art such as the Pearson correlation coefficient to obtain the correlation between the two component signals, and this application does not impose any special restrictions.

[0059] The independence of each component signal of the vibration signal is determined according to the correlation between different component signals of the vibration signal.

[0060] Preferably, as an embodiment of the present application, the correlations between all different component signals are normalized, and the normalized value of the correlations between different component signals is recorded as the normalized correlations between different component signals, the difference between the number 1 and the normalized correlations between different component signals is recorded as the difference between different component signals, and the sum of the differences between any component signal and all other component signals is recorded as the independence of any component signal.

[0061] Specifically, this embodiment selects the minimum-maximum normalization method to normalize the correlation between different component signals. As other implementation methods, on the basis of the purpose of normalizing the data, the implementer may adopt other methods in the prior art, such as the Z-Score standard normalization method, to normalize the data, and this application does not make any special restrictions; this embodiment uses the method of the difference between the number 1 and the normalized value of the data to determine the difference between different component signals. On the premise of ensuring that the difference between different component signals is negatively correlated with the normalized correlation between different component signals, the implementer may adopt other methods in the prior art, such as the Z-Score standard normalization method, to normalize the data and determine the difference between different component signals, and this application does not make any special restrictions.

[0062] Specifically, the calculation formula for the independence of the component signals of the vibration signal is:

[0063]

[0064] In the formula, The vibration signal Independence of the component signals; The vibration signal component signals; The vibration signal component signals; represents a first preset constant; represents a second preset constant. In this embodiment, the value of the second preset constant is 1; The vibration signal The component signal and The normalized correlation of the component signals.

[0065] When The greater the correlation between the first component signal and all other component signals, the The more obvious the correlation between the first component signal and all other component signals, the The more likely the component signal is formed by the interaction of multiple faults, the more likely the component signal is formed by the interaction of multiple faults. The smaller the independence of the component signals.

[0066] According to the fluctuation degree and independence of the component signal of the vibration signal, the fault characteristic value of the component signal of the vibration signal is obtained. The fault characteristic value of the component signal of the vibration signal is positively correlated with the fluctuation degree and independence of the component signal of the vibration signal.

[0067] Preferably, as an embodiment of the present application, the product of the fluctuation degree and independence of the component signal of the vibration signal is recorded as the fault characteristic value of the component signal of the vibration signal.

[0068] Specifically, the calculation formula of the fault characteristic value of the component signal of the vibration signal is:

[0069]

[0070] In the formula, The vibration signal The fault characteristic value of the component signal; The vibration signal The volatility of the component signals; The vibration signal The independence of the component signals.

[0071] The fault characteristic value of the component signal is a significance evaluation of the composite fault characteristics presented by the component signal. When the volatility and independence of the component signal of the vibration signal are greater, the fault characteristic value of the component signal of the vibration signal is greater. At this time, the component signal is more likely to be a component signal with a composite fault.

[0072] At this point, the fault characteristic value of each component signal of the vibration signal is obtained. The fault characteristic value acquisition flow chart is as follows: Figure 2 shown.

[0073] Step S003, based on the amplitude values ​​of the vibration signal of the same component signal at different acquisition moments, determine the local energy of each component signal at any acquisition moment, and determine the signal fault degree of the component signal of the vibration signal based on the fault characteristic value of the component signal of the vibration signal and the difference in the local energy of the component signal of the vibration signal at two adjacent acquisition moments.

[0074] When a bearing has a complex fault, the energy of the vibration signal will change suddenly, that is, the energy of the vibration signal will increase significantly.

[0075] For any component signal of the vibration signal, any moment of vibration signal collection is recorded as the target moment. This embodiment is described by taking any moment as an example. Taking the amplitude of the vibration signal of any component signal at the target moment as the center, a local window of the target moment of any component signal with a length of the third preset constant is established, and the sum of the squares of the amplitudes of all vibration signals contained in the local window of any component signal at the target moment is recorded as the local energy of any component signal at the target moment. In this embodiment, the value of the third preset constant is 5.

[0076] The same method can be used to obtain the local energy of any component signal at any time when the vibration signal is collected. That is, for any time when the vibration signal is collected, each component signal of the vibration signal has a corresponding local energy.

[0077] The signal fault degree of the component signal of the vibration signal is determined according to the fault characteristic value of the component signal of the vibration signal and the difference of the local energy of the component signal of the vibration signal at two adjacent acquisition moments. The signal fault degree of the component signal of the vibration signal is positively correlated with the fault characteristic value of the component signal of the vibration signal and the difference of the local energy of the component signal of the vibration signal at two adjacent acquisition moments.

[0078] Preferably, as an embodiment of the present application, the absolute value of the difference in local energy between the component signals of the vibration signal at the target moment and the next adjacent moment of the target moment is recorded as the local energy difference of the component signals of the vibration signal at the target moment, the cumulative sum of the local energy differences of the component signals of the vibration signal at all moments is recorded as the energy difference change of the component signals of the vibration signal, and the product of the energy difference change of the component signals of the vibration signal and the fault characteristic value is recorded as the signal fault degree of the component signals of the vibration signal.

[0079] It can be understood that, when establishing a local window of a component signal at a certain moment, if a local window with a length of the third preset constant cannot be established, or when calculating the local energy difference of the component signal at a certain moment, the local energy of the next adjacent moment of the certain moment does not exist, the moment when the above situation occurs will not be included in the process of calculating the signal fault degree of the component signal.

[0080] In actual application, as another implementation method, the implementer may also record the sum of the energy difference change of the component signal of the vibration signal and the fault characteristic value as the signal fault degree of the component signal of the vibration signal.

[0081] The signal fault degree of the component signal is a comprehensive evaluation of the significance of the composite fault presented by the component signal. When the difference between the fault characteristic value of the component signal of the vibration signal and the local energy of the component signal of the vibration signal at two adjacent acquisition moments is greater, the signal fault degree of the component signal is greater, and the significance of the composite fault presented by the component signal is greater.

[0082] At this point, the signal failure degree of the component signal of the vibration signal is obtained.

[0083] Step S004, determining whether there is an abnormal signal according to the signal fault degree of all component signals of the vibration signal, and obtaining the reactor fault diagnosis result.

[0084] The signal fault degree of all component signals of the vibration signal is detected abnormally. In this embodiment, the HBOS abnormal point detection algorithm is selected to detect the signal fault degree of all component signals of the vibration signal abnormally, and the abnormal scores of the signal fault degree of all component signals are obtained.

[0085] Among them, the number of parameter division intervals of the HBOS outlier detection algorithm is , represents the number of component signals into which the vibration signal is decomposed, i.e., a first preset constant; represents a rounding function, which is used to round off the value in the brackets. Using the HBOS outlier detection algorithm to perform anomaly detection on data is a well-known technology and will not be described in detail.

[0086] Setting anomaly thresholds In this embodiment, the abnormal threshold The value is 0.8. The anomaly score is greater than or equal to the anomaly threshold The component signal corresponding to the signal fault degree is marked as an abnormal signal.

[0087] In actual application, as other implementation methods, on the basis of achieving the purpose of abnormal detection of signal fault degrees of all component signals of the vibration signal, the implementer may adopt other methods in the prior art such as the Bayesian change point detection algorithm for abnormal detection, and this application does not impose any special restrictions.

[0088] When an abnormal signal is identified, it is determined that the reactor is faulty; when no abnormal signal is identified, it is determined that the reactor is operating normally.

[0089] At this point, intelligent diagnosis of reactor faults has been achieved.

[0090] Based on the same inventive concept as the above method, an embodiment of the present invention also provides an intelligent diagnosis system for reactor faults for acidic reactions, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor implements the steps of any one of the above-mentioned intelligent diagnosis methods for reactor faults for acidic reactions when executing the computer program.

[0091] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. An intelligent diagnosis method for reactor faults in acidic reactions, characterized in that: The method comprises the following steps: Collect vibration signals of the reactor's bearings at different collection times; Decomposing the vibration signal into component signals of a preset number of frequency bands, determining the degree of fluctuation of the component signals of the vibration signal according to the difference between the component signals of the vibration signal and the fluctuation of the amplitude contained in the vibration signal, determining the independence of each component signal of the vibration signal according to the correlation between different component signals of the vibration signal, and obtaining the fault characteristic value of the component signal of the vibration signal according to the degree of fluctuation and independence of the component signals of the vibration signal; According to the amplitude values ​​of the vibration signal of the same component signal at different acquisition moments, the local energy of each component signal at any acquisition moment is determined, and according to the fault characteristic value of the component signal of the vibration signal and the difference in the local energy of the component signal of the vibration signal at two adjacent acquisition moments, the signal fault degree of the component signal of the vibration signal is determined; Determine whether there is an abnormal signal based on the signal fault degree of all component signals of the vibration signal, and obtain the reactor fault diagnosis result; The specific method for obtaining the fluctuation degree of the component signal of the vibration signal is as follows: The information entropy of the amplitude contained in the component signal of the vibration signal is recorded as the first information entropy of the component signal, the information entropy of the amplitude contained in the vibration signal is recorded as the second information entropy of all component signals of the vibration signal, and the ratio of the first information entropy to the second information entropy of the component signal is recorded as the fluctuation degree of the component signal.

2. The intelligent diagnosis method for reactor faults in acidic reactions according to claim 1, characterized in that: The specific method of determining the independence of each component signal of the vibration signal according to the correlation between different component signals of the vibration signal includes: Calculate the correlation between different component signals of the vibration signal, and record the normalized value of the correlation between the different component signals as the normalized correlation between the different component signals; The independence of the component signals is obtained based on the normalized correlation between all different component signals.

3. The intelligent diagnosis method for reactor faults in acidic reactions according to claim 2 is characterized in that: The specific method of obtaining the independence of the component signals according to the normalized correlation between all different component signals is as follows: Determining the difference between the different component signals according to the normalized correlation between the different component signals, wherein the difference between the different component signals is negatively correlated with the normalized correlation between the different component signals; The sum of the differences between any component signal and all other component signals is recorded as the independence of the component signal.

4. The intelligent diagnosis method for reactor faults in acidic reactions according to claim 1, characterized in that: The specific method for obtaining the fault characteristic value of the component signal of the vibration signal is as follows: The product of the fluctuation degree and independence of the component signal of the vibration signal is recorded as the fault characteristic value of the component signal of the vibration signal.

5. The intelligent diagnosis method for reactor faults in acidic reactions according to claim 1, characterized in that: The method of determining the local energy of each component signal at any acquisition time according to the amplitude values ​​of the vibration signal of the same component signal at different acquisition times includes the following specific methods: Record any moment of vibration signal collection as the target moment; For the same component signal, the sum of the squares of the amplitudes of the vibration signals at the time when the third preset constant number of vibration signals centered at the target time is collected is recorded as the local energy of the same component signal at the target time.

6. The intelligent diagnosis method for reactor faults in acidic reactions according to claim 5, characterized in that: The signal fault degree of the component signal of the vibration signal includes the following specific methods: The absolute value of the difference between the local energy of the component signal of the vibration signal at the target time and the next adjacent time of the target time is recorded as the local energy difference of the component signal of the vibration signal at the target time; The signal fault degree of the component signal is obtained according to the fault characteristic value of the component signal and the local energy difference of the component signal at all times.

7. The intelligent diagnosis method for reactor faults in acidic reactions according to claim 6, characterized in that: The method of obtaining the signal fault degree of the component signal according to the fault characteristic value of the component signal and the local energy difference of the component signal at all times includes: The cumulative sum of the local energy differences of the component signals of the vibration signal at all times is recorded as the energy difference change of the component signals of the vibration signal; The product of the energy difference change of the component signal of the vibration signal and the fault characteristic value is recorded as the signal fault degree of the component signal of the vibration signal.

8. The intelligent diagnosis method for reactor faults in acidic reactions according to claim 1, characterized in that: The specific method of determining whether there is an abnormal signal according to the signal fault degree of all component signals of the vibration signal and obtaining the reactor fault diagnosis result includes: Perform anomaly detection on the signal fault degrees of all component signals of the vibration signal, obtain anomaly scores of the signal fault degrees of all component signals, and mark the component signals corresponding to the signal fault degrees whose anomaly scores are greater than or equal to an anomaly threshold as abnormal signals; When an abnormal signal is identified, it is determined that the reactor is faulty; when no abnormal signal is identified, it is determined that the reactor is operating normally.

9. An intelligent fault diagnosis system for reactors in acidic reactions, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 8 are implemented.

Images