A dry powder extinguishing agent production line working condition data processing method and system

Abstract

The present application belongs to the technical field of data processing, and particularly relates to a dry powder extinguishing agent production line working condition data processing method and system, which comprises the following steps: collecting a frictional electrostatic signal on a pneumatic conveying pipeline and extracting a high-frequency component; calculating a flow decoupling electrification index, eliminating the influence of flow changes by using the ratio relationship between energy and fluctuation; calculating dielectric relaxation entropy, quantifying the escape degree of electric charge in the transmission process; constructing a two-dimensional state space, identifying false aggregation or raw material failure faults according to the distribution of the index and entropy value, and adjusting the temperature of a reaction kettle or the feeding speed. The present application can effectively distinguish flow fluctuation and membrane quality change, solves the problem of being unable to perceive false aggregation online, and realizes online accurate monitoring and closed-loop control of the hydrophobic membrane integrity of the dry powder extinguishing agent.

Description

Technical Field

[0001] This invention relates to the field of data processing technology. More specifically, this invention relates to a method and system for processing operating data of a dry powder fire extinguishing agent production line. Background Technology

[0002] As the most widely used fire extinguishing medium in the current fire protection field, ABC dry powder fire extinguishing agent usually has ammonium dihydrogen phosphate as its core component. Because ammonium dihydrogen phosphate has extremely strong hygroscopicity, it is very easy to absorb moisture and hydrolyze in the natural environment, which leads to powder agglomeration and loss of fluidity. Therefore, in the industrial production process of dry powder fire extinguishing agent, surface hydrophobic modification is a key process that determines product quality.

[0003] This process typically involves uniformly spraying atomized methyl hydrogen silicone oil onto the surface of powder particles in a high-temperature reactor. Under the influence of a catalyst and a thermal field, a dehydrogenation crosslinking polymerization reaction occurs, thereby forming a dense polysiloxane hydrophobic film on the surface of micron-sized particles in situ. The integrity and density of this film directly determine the water repellency level, electrical insulation properties, and long-term storage stability of the finished product.

[0004] However, in existing industrial production practices, there are significant technical bottlenecks in the online monitoring and control of hydrophobic film formation quality. Traditional production control mainly relies on DCS systems to set and monitor macroscopic process parameters such as reactor jacket temperature, stirring speed, and reaction time. However, due to batch differences in upstream raw material ore, there are unavoidable fluctuations in particle size distribution, initial moisture content, and specific surface area. Under fixed process parameters, these fluctuations can lead to huge deviations in film formation, such as pseudo-polymerization where silicone oil has been sprayed but not fully cured, or incomplete coating due to increased specific surface area. DCS systems based on macroscopic parameter monitoring are completely unable to detect such changes in microscopic physicochemical states, resulting in an extremely high false alarm rate, making it impossible to implement online monitoring systems in actual closed-loop control. Summary of the Invention

[0005] To address the technical problems in the prior art, such as the inability to distinguish between flow fluctuations and membrane quality changes, which leads to the inability to detect false polymerization or incomplete coating of dry powder fire extinguishing agents online, resulting in a high false alarm rate and the inability to perform closed-loop control, the present invention provides solutions in the following aspects.

[0006] In a first aspect, the present invention provides a method for processing operating condition data of a dry powder fire extinguishing agent production line, comprising: collecting raw time-series data from upstream and downstream sensors on a pneumatic conveying pipeline between the outlet of the siliconization reactor and the finished product warehouse, and decomposing and reconstructing the raw time-series data to obtain triboelectric dynamic signals of the upstream and downstream sensors respectively; calculating the root mean square value, standard deviation, and center frequency of the power spectrum of the signal based on the triboelectric dynamic signal of the upstream sensor, and then calculating the flow decoupling electrification index; calculating the maximum correlation coefficient of the cross-correlation function based on the triboelectric dynamic signals of the upstream and downstream sensors, and calculating the zero point of the autocorrelation function as the energy of the upstream and downstream sensor signals; calculating the absolute value of the energy difference between the upstream and downstream sensor signals as the absolute value of transmission energy loss; calculating the dielectric relaxation entropy based on the maximum correlation coefficient, energy, and absolute value of transmission energy loss; constructing a two-dimensional state space based on the flow decoupling electrification index and the dielectric relaxation entropy, identifying the fault type of the production line operating condition based on the two-dimensional state space, and performing closed-loop control of the production line equipment according to the fault type.

[0007] By collecting and reconstructing dynamic signals of triboelectric static electricity and calculating the flow decoupling electrification index and dielectric relaxation entropy, this invention can extract features characterizing the microscopic surface state from complex operating noise: the flow decoupling electrification index eliminates the influence of macroscopic flow fluctuations on the signal and simply reflects the triboelectric charging capability of the particle surface; the dielectric relaxation entropy quantifies the degree of charge leakage during transmission and directly reflects the insulation integrity of the hydrophobic film; the two-dimensional state space constructed based on these two indicators can accurately distinguish between normal operating conditions, pseudo-polymerization faults, and raw material failure faults, thereby guiding the production line to carry out targeted closed-loop control, improving product quality consistency and production efficiency.

[0008] Preferably, the original time-series data is decomposed and reconstructed to obtain the triboelectric dynamic signals of the upstream sensor and the downstream sensor, respectively, including: decomposing the original time-series data of the upstream sensor and the downstream sensor using an ensemble empirical mode decomposition algorithm; discarding the first two low-frequency components obtained from the decomposition and retaining the high-frequency components; and reconstructing the triboelectric dynamic signals of the upstream sensor and the downstream sensor using the retained high-frequency components.

[0009] Preferably, the formula for calculating the flow decoupling power generation index is: In the formula, For a moment The flow decoupling power generation index; For a moment The root mean square value of the signal within the sliding window; For a moment The standard deviation of the signal within the sliding window; To prevent zero-noise substrate; For a moment The center frequency of the power spectrum within the sliding window; The reference frequency; It is the natural logarithm function.

[0010] This invention constructs a flow decoupling electrification index that includes the ratio of energy and fluctuation terms. By utilizing the division relationship between the root mean square value of the signal representing total energy and the standard deviation of the signal representing the intensity of flow field fluctuations, the influence of flow changes is automatically offset. At the same time, a logarithmic frequency domain correction term is introduced. By utilizing the difference in frequency response between well-silicified powder and coarse particles, the sensitivity of the index to the microscopic surface state is further enhanced, thereby accurately characterizing the film quality even under flow fluctuations.

[0011] Preferably, the method for obtaining the zero-noise substrate and reference frequency is as follows: Perform a self-calibration mode: start the induced draft fan but do not open the discharge valve; when only air flows in the pipeline, collect the signal from the upstream sensor and calculate its standard deviation as the zero-noise substrate; when standard dry powder is being transported in the pipeline, perform power spectrum analysis on the signal from the upstream sensor and extract the main frequency peak as the reference frequency. .

[0012] Preferably, the calculation of the maximum correlation coefficient of the cross-correlation function includes: the cutoff time. The upstream and downstream sensor signals within the sliding window; calculate the time delay of the upstream and downstream sensor signals. Cross-correlation function under : In the formula, The size of the sliding window; For upstream sensors at time The dynamic signal of triboelectric static electricity; For upstream sensors at time The dynamic signal of triboelectric static electricity; in time delay Search within the range to find the maximum correlation coefficient; This is the theoretical maximum transmission time. , The distance between the upstream and downstream sensors. This indicates the minimum flow rate.

[0013] By calculating the cross-correlation coefficient, we can analyze the changes in the signal from the waveform similarity in the time domain. The maximum correlation coefficient reflects the degree of waveform distortion, providing data support for the subsequent calculation of dielectric relaxation entropy.

[0014] Preferably, the dielectric relaxation entropy is calculated as follows: In the formula, For a moment The dielectric relaxation entropy; For a moment The maximum correlation coefficient of the cross-correlation function; For the upstream sensor signal at time Energy; For downstream sensor signals at time Energy; Let be the absolute value of transmission energy loss, and ; The reference energy loss constant; It is a natural exponential function.

[0015] This invention constructs a comprehensive index called dielectric relaxation entropy, which combines waveform similarity with energy retention factor: when the film layer is damaged or absorbs moisture, the rapid discharge of charge leads to waveform distortion and large energy loss, resulting in a significant increase in dielectric relaxation entropy; this index directly quantifies the degree of insulation defects on the powder surface and can sensitively detect microscopic damage or conductive channels in the film layer.

[0016] Preferably, the method for obtaining the reference energy loss constant is as follows: the signals from the upstream sensor and the downstream sensor collected during the self-test calibration mode are divided into segments, the energy difference of each segment is calculated, and the average value of the energy difference of all segments is used as the reference energy loss constant.

[0017] Preferably, the fault type of the production line is identified based on the two-dimensional state space, and the production line equipment is controlled in a closed loop according to the fault type, including: if the dielectric relaxation entropy is less than a preset dielectric relaxation threshold and the flow decoupling electrification index is greater than a preset electrification index threshold, the operating condition is determined to be normal, and the current process parameters are maintained; if the flow decoupling electrification index is greater than a preset electrification index threshold, but the dielectric relaxation entropy is greater than a preset dielectric relaxation threshold, the fault type is determined to be pseudo-polymerization, and an instruction to increase the reactor jacket temperature setpoint and reduce the speed of the discharge star valve is sent to the production line control system; otherwise, the fault type is determined to be raw material failure, the flow feedback of the silicone oil metering pump is checked, if the flow is low, the pump frequency is increased, and if the flow is normal, an audible and visual alarm for raw material moisture is triggered.

[0018] This invention, through the division of a two-dimensional state space, can clearly identify two typical production faults: pseudo-polymerization and raw material failure, and provides targeted closed-loop control strategies, reducing reliance on operator experience and improving the automation level of the production process.

[0019] Preferably, the method further includes: packaging the timestamp, batch number, flow decoupling charge index, dielectric relaxation entropy and judgment result, encrypting and storing them in a local audit log database, and generating a quality fingerprint code based on the data in the audit log database when the product leaves the factory.

[0020] In a second aspect, the present invention provides a dry powder fire extinguishing agent production line condition data processing system, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned dry powder fire extinguishing agent production line condition data processing method is implemented.

[0021] By adopting the above technical solution, a computer program is generated from the above-mentioned method for processing the operating data of a dry powder fire extinguishing agent production line, and stored in a memory for loading and execution by a processor. Terminal equipment is then manufactured based on the memory and processor for convenient use.

[0022] The beneficial effects of this invention are as follows:

[0023] By collecting and reconstructing dynamic signals of triboelectric static electricity and calculating the flow decoupling electrification index and dielectric relaxation entropy, this invention can extract features characterizing the microscopic surface state from complex operating noise: the flow decoupling electrification index eliminates the influence of macroscopic flow fluctuations on the signal and simply reflects the triboelectric charging capability of the particle surface; the dielectric relaxation entropy quantifies the degree of charge leakage during transmission and directly reflects the insulation integrity of the hydrophobic film; the two-dimensional state space constructed based on these two indicators can accurately distinguish between normal operating conditions, pseudo-polymerization faults, and raw material failure faults, thereby guiding the production line to carry out targeted closed-loop control, improving product quality consistency and production efficiency. Attached Figure Description

[0024] Figure 1 This is a flowchart illustrating a method for processing operating data of a dry powder fire extinguishing agent production line according to the present invention.

[0025] Figure 2 This is a schematic diagram showing a comparison between the traditional RMS index and the flow decoupling power generation index of the present invention;

[0026] Figure 3 This is a schematic diagram illustrating fault diagnosis based on a two-dimensional state space. Detailed Implementation

[0027] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0028] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0029] This invention discloses a method for processing operating data of a dry powder fire extinguishing agent production line, referring to... Figure 1 This includes steps S1-S4:

[0030] S1: Collect the raw time series data of the upstream and downstream sensors on the pneumatic conveying pipeline between the outlet of the siliconization reactor and the finished product warehouse, and decompose and reconstruct the raw time series data to obtain the triboelectric dynamic signals of the upstream and downstream sensors respectively.

[0031] It should be noted that in order to ensure the accuracy of physical constants in subsequent calculations, the current system's baseline state must be obtained; at the same time, there are power frequency and low-frequency vibrations in the pipeline on site, and the directly collected signals contain a lot of noise and cannot be directly used for microscopic feature extraction.

[0032] Specifically, two non-contact annular electrostatic sensors, labeled upstream sensor A and downstream sensor B, are installed along the powder flow direction on the pneumatic conveying pipeline between the outlet of the siliconization reactor and the finished product silo. The distance between the two sensors is... It is 0.3 meters.

[0033] Furthermore, a 10-minute self-calibration mode was executed: the induced draft fan was started but the discharge valve was not opened, only pure air flowed in the pipeline, the background noise signal of upstream sensor A was collected, its standard deviation was calculated, and it was set as the zero-noise baseline. Subsequently, a trial run was conducted using standard ABC dry powder, known to have fully qualified water repellency. Power spectrum analysis was performed on the signal from upstream sensor A, and its dominant frequency peak was extracted and stored as the reference frequency. .

[0034] Furthermore, the continuous signals from sensors A and B collected over these 10 minutes are divided into numerous 1-second segments. The zero point of the autocorrelation function for each segment is calculated as the energy. For each segment, the average of the squares of the signal values ​​from sensor A is calculated and recorded as the energy of the segment for sensor A. Similarly, the average of the squares of the signal values ​​from sensor B is calculated and recorded as the energy of the segment for sensor B. The energy difference between each segment of sensor A and the corresponding segment of sensor B is calculated, and the average of the energy differences across all segments is taken as the reference energy loss constant. .

[0035] It should be noted that the mean square value of the signal represents the total intensity of the electrostatic signal sensed by the sensor at that instant. During the process of powder flowing from upstream A to downstream B, charge will naturally leak to the pipe wall. Therefore, the signal energy at upstream A is usually slightly greater than the signal energy at downstream B. The energy difference represents the amount of natural charge loss of a standard qualified product during normal transmission. This difference is calibrated and used as a benchmark to measure whether charge loss is too rapid in subsequent production. If the loss far exceeds the benchmark energy loss constant... This indicates that the insulation of the film layer has deteriorated.

[0036] Furthermore, during the formal production process, sensor A and sensor B use a sampling rate... Synchronously acquire induced current signals to obtain raw time-series data. Due to power frequency interference from motor inverters and mechanical vibration noise from pipelines in industrial settings, directly using the raw signals would reduce computational accuracy. Therefore, the Ensemble Empirical Mode Decomposition (EEMD) algorithm is used to decompose the raw time-series data, discarding the first two low-frequency components representing macroscopic flow pattern fluctuations and retaining the high-frequency components representing microscopic particle collision and friction characteristics, thus reconstructing a pure triboelectric dynamic signal. and High-frequency components are mainly generated by collisions and friction between microscopic particles and the pipe wall, and between particles themselves, and best reflect the physical state of the particle surface; low-frequency components are mostly macroscopic dune flow characteristics and are therefore filtered out.

[0037] S2: Based on the dynamic signal of triboelectric static electricity from the upstream sensor, calculate the root mean square value of the signal, the standard deviation of the signal, and the center frequency of the power spectrum, and then calculate the flow decoupling electrification index.

[0038] It should be noted that in gas-solid two-phase flow, the effective value of the electrostatic signal is proportional to the total charge and is affected by both flow rate and film layer, while the standard deviation is proportional to the intensity of turbulent fluctuations and is only affected by flow rate. Existing technologies cannot distinguish whether signal fluctuations originate from changes in flow rate or changes in film layer quality. Therefore, this invention constructs a flow rate decoupling electrification index and utilizes the ratio of energy to fluctuation to automatically offset the influence of flow rate changes.

[0039] Specifically, based on the triboelectric dynamic signal from the upstream sensor By setting a sliding window and calculating the root mean square value of the signal, the standard deviation of the signal, and the center frequency of the power spectrum within the window, the formula for calculating the flow decoupling electrification index is:

[0040]

[0041] In the formula, For a moment The flow decoupling power generation index; For a moment The root mean square value of the signal within the sliding window represents the total energy; For a moment The standard deviation of the signal within the sliding window characterizes the intensity of flow field fluctuations; To prevent zero-noise basis, used to prevent the denominator from approaching zero and causing calculation divergence; For a moment The center frequency of the power spectrum within the sliding window; The reference frequency; It is the natural logarithm function.

[0042] Among them, the size of the sliding window This determines the temporal resolution and stability of feature extraction. If it is too small, the statistical features fluctuate greatly; if it is too large, transient changes cannot be captured. The value range is [0.5s, 2s]. In this embodiment, the size of the sliding window is... Set to 1 second.

[0043] In this calculation, the flow field fluctuation intensity is introduced as a normalization factor in the denominator. As the production line output increases, the numerator... With denominator The synchronous increase and the division of the two cancel each other out, thus eliminating the influence of macroscopic flow rate changes on the indicators; well-silicified powder has a smooth surface and a low coefficient of friction, causing a shift in the frequency components of high-frequency noise generated by collisions; while coarse, unsilicified particles have different frequency responses, leading to the introduction of a logarithmic frequency domain correction term. This further enhances the sensitivity to the microscopic surface state.

[0044] For example, a comparison chart of the traditional RMS index and the flow decoupling power generation index of the present invention is shown below. Figure 2 As shown in the figure, the five regions correspond to the calibration region, normal region, flow fluctuation region, pseudo-polymerization region, and raw material failure region, respectively. In the flow fluctuation region, the traditional RMS index spikes dramatically with increasing flow rate. In existing technologies, this would be misjudged as excellent film quality or system anomaly. However, the flow decoupling electrification index of this invention remains relatively stable in this region. This is because the flow decoupling electrification index utilizes the synchronous change characteristics of the numerator representing energy and the denominator representing turbulence fluctuation to cancel out flow interference and extract an electrification index that is only related to the surface properties of the material.

[0045] S3: Based on the triboelectric dynamic signals from the upstream and downstream sensors, calculate the maximum correlation coefficient of the cross-correlation function, the zero point of the autocorrelation function, and the absolute value of the transmission energy loss; calculate the dielectric relaxation entropy based on the maximum correlation coefficient, energy, and absolute value of the transmission energy loss.

[0046] It should be noted that the microscopic integrity of the siliconized film determines its charge retention capability. If the film is damaged or conductive, the charge will rapidly leak to the tube wall during the process of the particles flowing from upstream sensor A to downstream sensor B, i.e., dielectric relaxation will occur, resulting in attenuation of downstream signal energy and waveform distortion. Therefore, this invention quantifies the degree of charge escape during the transmission process by constructing dielectric relaxation entropy.

[0047] Specifically, at any given moment The system synchronously intercepts signals from upstream sensors. and downstream sensor signals In the interval The window data fragments within are denoted as follows: and .

[0048] Furthermore, based on the captured window data fragments and Calculate the cross-correlation function :

[0049]

[0050] In the formula, For the upstream sensor signal and the downstream sensor signal in time delay The cross-correlation coefficients under the following conditions; The size of the sliding window; For upstream sensors at time The dynamic signal of triboelectric static electricity; For upstream sensors at time The dynamic signal of triboelectric static electricity.

[0051] Time delay Search within the range to find the maximum value, and record it as . , characterizing waveform similarity; This is the theoretical maximum transmission time. This is used to limit the search scope and reduce the amount of computation. The distance between upstream sensor A and downstream sensor B. This indicates the minimum flow rate.

[0052] Furthermore, calculate the window data fragments separately. and The zero point of the autocorrelation function is used as the reference point for the upstream and downstream sensor signals at time t. The energy; then calculate the absolute value of the difference between the two, as the time. absolute value of transmission energy loss .

[0053] Finally, calculate the time. The dielectric relaxation entropy is calculated using the following formula:

[0054]

[0055] In the formula, For a moment The dielectric relaxation entropy; For a moment The maximum correlation coefficient of the cross-correlation function characterizes the waveform similarity; For the upstream sensor signal at time Energy; For downstream sensor signals at time Energy; Let be the absolute value of transmission energy loss, and ; It serves as the benchmark energy loss constant, used as a standard for judging the rate of energy decay; It is a natural exponential function.

[0056] in, This indicates the maximum similarity between two waveforms. This is a normalization factor used to eliminate the influence of the absolute amplitude of the signal, so that the correlation coefficient is in the range of [0,1]. It is the energy retention factor; when the film layer is intact and dense, there is no charge leakage, and the waveforms are highly similar, that is... High, energy loss As the exponent approaches zero, the exponent term approaches 1, at which point the subtrahend term also approaches 1, and finally, the dielectric relaxation entropy... Approaching zero; when the film layer is damaged or absorbs moisture, the charge is rapidly released, causing waveform distortion, i.e. Low energy loss When the exponential term approaches zero, the subtrahend term decreases, and finally, the dielectric relaxation entropy... Approaching 1; therefore, the dielectric relaxation entropy It directly quantifies the degree of insulation defects on the powder surface; the lower the value, the better the film quality.

[0057] S4: Construct a two-dimensional state space based on the flow decoupling power index and dielectric relaxation entropy, identify the fault types of the production line based on the two-dimensional state space, and perform closed-loop control of the production line equipment according to the fault types.

[0058] It should be noted that a single indicator is insufficient to fully reflect complex production failures. By establishing a two-dimensional state space, different failure modes can be accurately identified, and process parameters can be adjusted accordingly, reducing reliance on operator experience.

[0059] Specifically, the dielectric relaxation entropy is established. Decoupling from flow rate and power generation index The two-dimensional state space is constructed, and a dielectric relaxation threshold is set. and the threshold of the charging index The system performs diagnostics and controls periodically.

[0060] (1) When and When the current condition is reached, it is determined to be a normal operating condition, that is, the current is strong and there is no leakage, and the current process parameters are maintained.

[0061] (2) When but When the failure mode is A, i.e., pseudo-polymerization, the silicone oil has been sprayed in place, but due to insufficient heat, it has not been fully cross-linked and cured, resulting in microscopic conductive channels in the film layer. The closed-loop control strategy is to send a command to the DCS system to increase the temperature setpoint of the reactor jacket by 2 degrees Celsius, and to send a command to the frequency converter to reduce the speed of the discharge star valve by 5% to prolong the residence time of the material in the reactor.

[0062] (3) When When the material fails, it is determined to be fault mode B, i.e., the raw material has failed. At this time, the energization index is extremely low, indicating that the silicone oil coating rate is seriously insufficient or the moisture content of the raw material is too high. The closed-loop control strategy is to check the flow feedback of the silicone oil metering pump. If the flow rate is low, the pump frequency is increased. If the flow rate is normal, the audible and visual alarm for the raw material being damp is triggered.

[0063] Among them, dielectric relaxation threshold and the threshold of the charging index The values ​​are empirical values ​​derived from statistical analysis of historical production data. In this embodiment, they are used to illustrate these values. Set to 0.8, It is set to 0.68; in other embodiments, implementers may adjust it based on the historical data distribution of the actual production line.

[0064] Finally, the timestamp, batch number, flow decoupling charge index, dielectric relaxation entropy, and judgment result are packaged, encrypted, and stored in a local, tamper-proof audit log database, and a quality fingerprint code is generated when the product leaves the factory.

[0065] It should be noted that, in order to meet the stringent traceability requirements of fire protection products, the system packages the above features and judgment results, and the data is encrypted and stored in a local, tamper-proof audit log database. When the product leaves the factory, the quality fingerprint code generated by the log will be printed on the product certificate of conformity, serving as a digital certificate of the batch's online full inspection of water repellency.

[0066] For example, a schematic diagram of fault diagnosis based on two-dimensional state space is shown below. Figure 3As shown in the figure, data points belonging to different operating conditions are divided into three different regions. Although the original signal amplitudes of data points belonging to the normal region and the flow fluctuation region differ greatly, they all fall within the normal operating condition region in this state space, proving the robustness of the invention. Data points belonging to the pseudo-polymerization region have a high flow decoupling electrostatic index value because the silicone oil is properly sprayed and has strong electrostatic charge. However, leakage current is caused by the incomplete curing of the film layer, resulting in energy loss. The increase leads to a significant increase in the dielectric relaxation entropy, which shifts to the right. This is a hidden fault that cannot be identified by simply looking at the traditional RMS index in existing technologies. Data points belonging to the raw material failure zone cannot be energized at all, and the value of the flow decoupling energization index is extremely low, falling directly into the lower region and being judged as a raw material problem. In summary, the fault diagnosis diagram intuitively shows the closed-loop control logic of the present invention. The system can accurately issue temperature adjustment or pump adjustment commands according to the quadrant where the data point is located.

[0067] This invention also discloses a dry powder fire extinguishing agent production line condition data processing system, including a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, a dry powder fire extinguishing agent production line condition data processing method according to the present invention is implemented.

[0068] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.

Claims

1. A method for processing operating condition data of a dry powder fire extinguishing agent production line, characterized in that, include: The raw time-series data of upstream and downstream sensors on the pneumatic conveying pipeline between the outlet of the siliconization reactor and the finished product warehouse are collected, and the raw time-series data are decomposed and reconstructed to obtain the triboelectric dynamic signals of the upstream and downstream sensors respectively. Based on the triboelectric dynamic signal from the upstream sensor, the root mean square value, standard deviation, and center frequency of the power spectrum are calculated, and then the flow decoupling electrification index is calculated. Based on the triboelectric dynamic signals of the upstream and downstream sensors, the maximum correlation coefficient of the cross-correlation function is calculated, and the zero point of the autocorrelation function is calculated as the energy of the upstream and downstream sensor signals. Calculate the absolute value of the energy difference between the upstream sensor signal and the downstream sensor signal as the absolute value of transmission energy loss; calculate the dielectric relaxation entropy based on the maximum correlation coefficient, energy, and absolute value of transmission energy loss. A two-dimensional state space is constructed based on the flow decoupling power-up index and the dielectric relaxation entropy. The production line fault type is identified based on the two-dimensional state space, and the production line equipment is controlled in a closed loop according to the fault type. The formula for calculating the flow decoupling electrification index is: In the formula, For a moment The flow decoupling power generation index; For a moment The root mean square value of the signal within the sliding window; For a moment The standard deviation of the signal within the sliding window; To prevent zero-noise substrate; For a moment The center frequency of the power spectrum within the sliding window; The reference frequency; It is the natural logarithm function; The formula for calculating dielectric relaxation entropy is: In the formula, For a moment The dielectric relaxation entropy; For a moment The maximum correlation coefficient of the cross-correlation function; For the upstream sensor signal at time Energy; For downstream sensor signals at time Energy; Let be the absolute value of transmission energy loss, and ; The reference energy loss constant; It is a natural exponential function.

2. The method for processing operating data of a dry powder fire extinguishing agent production line according to claim 1, characterized in that, The original time-series data is decomposed and reconstructed to obtain the triboelectric dynamic signals of the upstream and downstream sensors, respectively, including: The raw time-series data of the upstream and downstream sensors are decomposed using an ensemble empirical mode decomposition algorithm. Discard the first two low-frequency components obtained from the decomposition and retain the high-frequency components; The triboelectric dynamic signal of the upstream sensor and the triboelectric dynamic signal of the downstream sensor are reconstructed using the retained high-frequency components.

3. The method for processing operating data of a dry powder fire extinguishing agent production line according to claim 1, characterized in that, The method for obtaining the zero-noise substrate and the reference frequency is as follows: Execute self-calibration mode: Start the induced draft fan but do not open the discharge valve. When only air flows in the pipeline, collect the signal from the upstream sensor and calculate its standard deviation as a zero-noise baseline. When standard dry powder is transported in the pipeline, perform power spectrum analysis on the signal from the upstream sensor and extract the main frequency peak as the reference frequency. .

4. The method for processing operating data of a dry powder fire extinguishing agent production line according to claim 1, characterized in that, The calculation of the maximum correlation coefficient of the cross-correlation function includes: Intercepting time The upstream and downstream sensor signals within the sliding window; calculate the time delay of the upstream and downstream sensor signals. Cross-correlation function under : In the formula, The size of the sliding window; For upstream sensors at time The dynamic signal of triboelectric static electricity; For upstream sensors at time The dynamic signal of triboelectric static electricity; Time delay Search within the range to find the maximum correlation coefficient; This is the theoretical maximum transmission time. , The distance between the upstream and downstream sensors. This indicates the minimum flow rate.

5. The method for processing operating data of a dry powder fire extinguishing agent production line according to claim 4, characterized in that, The method for obtaining the reference energy loss constant is as follows: The signals from the upstream and downstream sensors collected during the self-calibration mode are segmented into segments, the energy difference of each segment is calculated, and the average of the energy differences of all segments is used as the reference energy loss constant.

6. The method for processing operating data of a dry powder fire extinguishing agent production line according to claim 1, characterized in that, Based on the two-dimensional state space, the production line operating condition fault types are identified, and closed-loop control of the production line equipment is performed according to the fault types, including: If the dielectric relaxation entropy is less than the preset dielectric relaxation threshold and the current decoupling charge index is greater than the preset charge index threshold, the operating condition is determined to be normal, and the current process parameters are maintained. If the flow decoupling electrification index is greater than the preset electrification index threshold, but the dielectric relaxation entropy is greater than the preset dielectric relaxation threshold, the fault type is determined to be pseudo-polymerization, and an instruction to increase the reactor jacket temperature setting value and reduce the discharge star valve speed is sent to the production line control system. Otherwise, determine the fault type as raw material failure, check the flow feedback of the silicone oil metering pump, if the flow is low, increase the pump frequency, if the flow is normal, trigger the raw material moisture alarm (audio and visual).

7. The method for processing operating data of a dry powder fire extinguishing agent production line according to claim 1, characterized in that, The method further includes: The timestamp, batch number, flow decoupling charge index, dielectric relaxation entropy, and judgment result are packaged, encrypted, and stored in the local audit log database. A quality fingerprint code is generated based on the data in the audit log database when the product leaves the factory.

8. A data processing system for the operating conditions of a dry powder fire extinguishing agent production line, characterized in that, include: A processor and a memory, wherein the memory stores computer program instructions that, when executed by the processor, implement a method for processing operating data of a dry powder fire extinguishing agent production line according to any one of claims 1-7.

Images