CN-121999592-A - Monitoring and early warning method and system for aluminum ash treatment waste gas of gas sensor

CN121999592ACN 121999592 ACN121999592 ACN 121999592ACN-121999592-A

Abstract

The invention discloses an aluminum ash treatment waste gas monitoring and early warning method and system of a gas sensor, and belongs to the technical field of aluminum ash treatment waste gas monitoring. The method comprises the steps of firstly obtaining multi-dimensional targeting data of the aluminum ash treatment whole flow waste gas, then completing feature extraction and redundancy elimination through a first analysis model to obtain a waste gas state core parameter set, inputting the waste gas state core parameter set into a second optimization model to obtain a pollutant accurate quantization parameter set, generating a multi-stage early warning signal and a process real-time adjustment instruction through a third early warning regulation model, and finally pushing the signal and issuing the instruction through a 5G and edge calculation node to realize real-time monitoring, dynamic early warning and closed-loop regulation of the waste gas. According to the invention, the three-core model is used for realizing the accurate extraction of the characteristics of the waste gas, the accurate quantification of the concentration of pollutants, the dynamic early warning and the accurate regulation, a closed-loop system is constructed by relying on 5G and edge calculation, the reliable monitoring, the timely early warning and the accurate regulation are ensured, the overall efficiency of waste gas treatment is improved, and the power assisting environment-friendly standard reaches the standard and the production is stable.

Inventors

WAN ZHIYONG
ZHANG JIN
QIAN JIE
YU XIANWEN

Assignees

湖北宇宸新材料科技有限公司

Dates

Publication Date: 20260508
Application Date: 20260210

Claims (10)

1. The method for monitoring and early warning the aluminum ash treated waste gas of the gas sensor is characterized by comprising the following steps of: Acquiring multi-dimensional targeting data of aluminum ash treatment whole-flow waste gas, wherein the targeting data comprises first core data and second core data; Performing feature extraction and redundancy elimination on the first core data and the second core data through a first analysis model to obtain a core parameter set of an exhaust gas state, wherein the first analysis model is constructed based on a time-space association mechanism of the physical and chemical properties of pollutants and the dynamic adaptation of multi-source signals; inputting the core parameter set of the waste gas state into a second optimization model to obtain a precise quantization parameter set of pollutants, wherein the second optimization model is constructed based on a deep confidence network and a least square support vector machine; Inputting the precise quantitative parameter set of the pollutants into a third early warning regulation model, generating a multi-stage early warning signal and a process real-time adjustment instruction, wherein the third early warning regulation model is constructed based on a fuzzy PID control logic and a sliding window dynamic threshold algorithm; the 5G and edge computing nodes push multi-stage early warning signals to the monitoring terminal, and the process real-time adjustment instruction is synchronously issued to the executing mechanism of the aluminum ash treatment device, so that the real-time monitoring, dynamic early warning and closed-loop regulation and control of the waste gas are realized.
2. The method for monitoring and early warning aluminum ash treatment waste gas of a gas sensor according to claim 1 is characterized in that the target data comprise first core data and second core data, the first core data are spectrum absorption signals, electrochemical response signals and concentration transient fluctuation data of fluoride, nitrogen oxide and particulate matters in waste gas, the second core data are treatment scene temperature and humidity gradual change data, air pressure steady value, air flow speed vector data and equipment operation load time sequence data, the concentration transient fluctuation data are quantized through sliding standard deviation, and the equipment operation load time sequence data are coupling data of real-time power and feeding rate of an aluminum ash treatment furnace.
3. The method for monitoring and early warning aluminum ash treated waste gas of a gas sensor according to claim 1, wherein the dynamic adaptation time-space association mechanism of the first analytical model is realized through a time-space weight matrix, signal association reference weight is determined based on pollution physical characteristics, and then the data time sequence correlation and the spatial distribution characteristic are combined for dynamic adjustment, and a characteristic extraction and redundancy elimination process formula is as follows: ; Wherein, the Is the first The core parameters of the state of the exhaust gas, Is the first Class-one core data Dynamically adapting weights for the second core data, For the characteristic value of the corresponding data, For the covariance of this feature with other features, For the maximum value of the covariance of all features, For the first number of core data types, Is the second core data type number.
4. The method for monitoring and pre-warning aluminum ash treated exhaust gas of a gas sensor according to claim 3, wherein the dynamic adaptive weights are Based on the physical and chemical properties of the pollutants and the data credibility, the data credibility is quantified through the signal to noise ratio, the physical and chemical properties of the pollutants comprise the spectrum absorption section, electrochemical activity, solubility and reactivity of the pollutants, and the calculation of the dynamic adaptation weight is required to meet the normalization constraint of all data association weights.
5. The method for monitoring and early warning aluminum ash treated waste gas of a gas sensor according to claim 1, wherein in the second optimization model, a deep belief network DBN comprises a plurality of RBM layers of a Boltzmann machine and an output layer, the number of neurons of each RBM layer is configured in a gradient decreasing manner, firstly, unsupervised pretraining is carried out on a core parameter set of a waste gas state, a contrast divergence algorithm is adopted in the pretraining process to minimize reconstruction errors, and then high-dimensional feature dimension reduction and reconstruction are realized through supervision fine tuning.
6. The method for monitoring and early warning the aluminum ash treatment waste gas of the gas sensor according to claim 5, wherein the second optimization model inputs the reconstruction characteristics output by the DBN into a least squares support vector machine LSSVM, and the accurate fitting of the concentration of the pollutants is realized through radial basis function mapping, and the fitting formula is as follows: ; Wherein, the Is the first The concentration of the pollutants is precisely quantified at the moment, Is Lagrange multiplier of LSSVM, For DBN output The features are reconstructed from the time of day, Is the first The reconstructed characteristics of the individual training samples are, As a parameter of the kernel function, As a result of the bias term, For the number of training samples.
7. The method for monitoring and early warning the aluminum ash treatment waste gas of the gas sensor according to claim 1, wherein the sliding window dynamic threshold of the third early warning regulation model is determined by the recent pollutant concentration statistical characteristic, the window length is adaptively set according to the aluminum ash treatment process period, and the dynamic early warning threshold formula is as follows: ; Wherein, the Is the first The moment dynamic early warning threshold value, As the mean value of the concentration within the sliding window, In order to obtain standard deviation of concentration in the sliding window, And the early warning sensitivity coefficient is positively and correspondingly adjusted along with the toxicity level of the pollutant.
8. The method for monitoring and early warning of aluminum ash treated waste gas of a gas sensor according to claim 1, wherein the fuzzy PID control logic of the third early warning regulation model comprises two input variables and three output variables, wherein the input variables are concentration deviations Rate of change of deviation The output variables are PID parameter correction amounts, namely proportional coefficient correction amounts, integral coefficient correction amounts and differential coefficient correction amounts, wherein the input variables and the output variables are configured with multi-gear fuzzy subsets, the parameter correction amounts are generated through a preset fuzzy rule base, and the fuzzy rule base is subjected to fuzzy solution through a gravity center method to obtain final correction values.
9. The method for monitoring and early warning aluminum ash treatment waste gas of the gas sensor according to claim 1 is characterized in that an executing mechanism of the aluminum ash treatment device comprises a waste gas purification spray pump, a ventilation regulating valve, a combustor power controller and a bag-type dust collector differential pressure regulator, generation of a process real-time adjustment instruction needs to be combined with environmental protection emission standards and equipment operation constraints, and instruction parameters need to be matched with a rated working range of the executing mechanism, so that equipment operation safety is ensured.
10. An aluminum ash treatment waste gas monitoring and early warning system of a gas sensor for realizing the method of any one of claims 1-9, which is characterized by comprising a data acquisition module, a model operation module, an early warning regulation module, a communication transmission module and an execution control module; The data acquisition module integrates a spectrum sensor, an electrochemical sensor, a temperature and humidity sensor, a barometric sensor, a wind speed sensor and a device load monitoring unit, and each sensor synchronously acquires data according to a uniform sampling frequency so as to meet the requirement of monitoring precision; The model operation module is deployed at the edge calculation node and supports the functions of online updating and offline training of model parameters; the communication transmission module adopts a 5G communication technology, so that the timeliness of data transmission rate and instruction issuing is ensured; The execution control module is internally provided with a driving circuit and a feedback acquisition unit, and can acquire the running state of the execution mechanism in real time and feed back the running state to the model operation module to form closed-loop control.

Description

Monitoring and early warning method and system for aluminum ash treatment waste gas of gas sensor Technical Field The invention relates to the technical field of monitoring of aluminum ash treatment waste gas, in particular to a method and a system for monitoring and early warning of aluminum ash treatment waste gas of a gas sensor. Background In the link of aluminum ash disposal and storage, aluminum ash is easy to release harmful gases such as ammonia gas, fluorine gas and the like, if the absorption and purification are not timely and the early warning is inaccurate, the atmospheric pollution can be caused, and the problems of personnel health risks and equipment corrosion can be possibly caused, so that the gas treatment and the safety early warning of an aluminum ash dangerous waste warehouse and a storage warehouse become the key requirements of the industry. Chinese patent (publication No. CN 220159634U) discloses an ammonia and fluorine gas absorbing and purifying device for an aluminum ash dangerous waste warehouse, which realizes harmful gas purification through an absorption tower structure, but focuses on static gas absorption treatment, and does not combine the dynamic characteristic design self-adaptive regulation function of aluminum ash released gas, so that when the concentration of the gas is suddenly increased due to the change of the aluminum ash stacking amount and the fluctuation of temperature and humidity, the purifying efficiency is easily affected, and the real-time matching high-efficiency purification is difficult to realize. Chinese patent (publication number: CN 216053274U) discloses a safety early warning system for an aluminum ash storage warehouse, which can monitor and trigger parameters such as gas concentration, temperature and humidity in the warehouse, but the early warning mechanism depends on a fixed threshold value, does not consider the difference of gas release rules in the aluminum ash storage period, does not form linkage control with a gas purifying device, can only realize the monitoring and early warning functions, cannot synchronously complete the closed loop response of early warning and purification, and is difficult to meet the double requirements of safety and environmental protection of the aluminum ash storage whole period. The prior art also has the problems that a gas purification and early warning system is disjointed, the purification parameters cannot be dynamically adapted to the working condition change, the early warning signals lack a grading treatment mechanism and the like, and the complexity and uncertainty of gas release in an aluminum ash dangerous waste warehouse and a storage warehouse are difficult to deal with, so that the technical scheme of the aluminum ash treatment waste gas monitoring and early warning method and the system of the gas sensor is needed. Disclosure of Invention Based on the technical problems, the application discloses an aluminum ash treatment waste gas monitoring and early warning method and system of a gas sensor, which specifically comprise the following steps: Acquiring multi-dimensional targeting data of aluminum ash treatment whole-flow waste gas, wherein the targeting data comprises first core data and second core data; Performing feature extraction and redundancy elimination on the first core data and the second core data through a first analysis model to obtain a core parameter set of an exhaust gas state, wherein the first analysis model is constructed based on a time-space association mechanism of the physical and chemical properties of pollutants and the dynamic adaptation of multi-source signals; inputting the core parameter set of the waste gas state into a second optimization model to obtain a precise quantization parameter set of pollutants, wherein the second optimization model is constructed based on a deep confidence network and a least square support vector machine; Inputting the precise quantitative parameter set of the pollutants into a third early warning regulation model, generating a multi-stage early warning signal and a process real-time adjustment instruction, wherein the third early warning regulation model is constructed based on a fuzzy PID control logic and a sliding window dynamic threshold algorithm; the 5G and edge computing nodes push multi-stage early warning signals to the monitoring terminal, and the process real-time adjustment instruction is synchronously issued to the executing mechanism of the aluminum ash treatment device, so that the real-time monitoring, dynamic early warning and closed-loop regulation and control of the waste gas are realized. Preferably, the targeting data comprises first core data and second core data, the first core data is spectrum absorption signals, electrochemical response signals and concentration transient fluctuation data of fluoride, nitrogen oxide and particulate matters in waste gas, the second core data is temperature and humidity gradual