Waste incineration flue gas denitration catalyst and preparation method thereof

Abstract

The invention relates to the field of intelligent preparation, and provides a waste incineration flue gas denitration catalyst and a preparation method thereof.The waste incineration flue gas denitration catalyst is characterized in that a time queue of slurry flow velocity values is collected through a flow velocity sensor, and segmentation and local time sequence correlation analysis are conducted on the slurry flow velocity values through a data processing and analysis algorithm based on a deep neural network; therefore, the recommended magnetic field intensity value of the next time point is automatically obtained according to the time-space significant aggregation between the local time sequence correlation characteristics of each slurry flow velocity value. Therefore, the magnetic field intensity can be automatically adjusted by quickly responding to the change of the flow velocity of the slurry, the iron oxide separation process is always in an optimal state, the need of manual intervention is reduced, and the automation level of the magnetic separation process is improved.

Description

Technical Field

[0001] The present application relates to the field of intelligent preparation, and more specifically, to a waste incineration flue gas denitrification catalyst and a preparation method thereof. Background Art

[0002] Waste incineration power generation is to put waste into an incinerator, where the combustible components are mixed with oxygen to burn and release heat, producing high-temperature flue gas and relatively stable residues. The heat generated during the incineration process is used to drive the unit to generate electricity. However, the waste incineration process will produce various pollutants, including nitrogen oxides (NOx), VOCs gas, dioxins, etc., which have serious negative effects on the environment and human health. Therefore, denitrification (removal of nitrogen oxides) of waste incineration flue gas has become an important environmental protection issue.

[0003] Patent CN114225936B proposes a waste incineration flue gas denitrification catalyst and its preparation method, which extracts α-Fe2O3 from red mud through magnetic separation, and prepares a mixed catalyst of α-Fe2O3 and γ-Fe2O3 through ball milling, screening, calcination and other steps to achieve low-temperature catalytic denitrification of waste incineration flue gas.

[0004] In the process of extracting iron oxide from red mud by magnetic separation, the control of magnetic field strength is very important. By precisely controlling the magnetic field strength, effective recovery of iron oxide can be achieved. However, in the above-mentioned patent, the traditional magnetic separation method is achieved by manually adjusting the excitation current or pole gap of the high-power magnetic separator. This method of manually adjusting the magnetic field strength is difficult to achieve high-precision control, especially when frequent adjustments are required, errors are prone to occur. Specifically, manual adjustment cannot respond to changes in fluid flow rate or slurry concentration in real time, resulting in unstable separation effect. In addition, frequent manual adjustments are not highly intelligent, which not only increases the workload of operators, but also easily leads to misoperation due to fatigue.

[0005] Therefore, an optimized method for preparing a waste incineration flue gas denitrification catalyst is desired. Summary of the invention

[0006] In view of the shortcomings of the prior art, the present application provides a waste incineration flue gas denitration catalyst and a preparation method thereof.

[0007] According to one aspect of the present application, a method for preparing a waste incineration flue gas denitration catalyst is provided, comprising: adding water to red mud and mixing uniformly to obtain uniformly mixed red mud; magnetically selecting iron oxide in the uniformly mixed red mud by a strong magnetic separator, and washing the screened iron oxide to obtain washed iron oxide; ball milling and sieving the washed iron oxide to obtain sieved iron oxide; calcining the sieved iron oxide in a reducing atmosphere to obtain the waste incineration flue gas denitration catalyst; wherein the magnetically selecting iron oxide in the uniformly mixed red mud by a strong magnetic separator comprises:

[0008] Obtaining a time queue of slurry flow rate values ​​collected by a flow rate sensor;

[0009] After the time queue of the slurry flow rate value is segmented into a sequence, the local time queue of each slurry flow rate value is sequence-encoded to obtain a time series of a local time series correlation feature vector of the slurry flow rate;

[0010] The time series of the local time series correlation feature vectors of the slurry flow velocity is subjected to node spatiotemporal significant aggregation to obtain a spatiotemporal significant aggregation representation of the slurry flow velocity, including: extracting the current local time series correlation feature vector of the slurry flow velocity from the time series of the local time series correlation feature vectors of the slurry flow velocity, and defining other local time series correlation feature vectors of the slurry flow velocity in the time series of the local time series correlation feature vectors of the slurry flow velocity as historical local time series correlation feature vectors of the slurry flow velocity to obtain a sequence of historical local time series correlation feature vectors of the slurry flow velocity; and performing feature significant aggregation based on energy transfer on the sequence of the historical local time series correlation feature vectors of the slurry flow velocity and the current local time series correlation feature vector of the slurry flow velocity to obtain the spatiotemporal significant aggregation representation of the slurry flow velocity;

[0011] Based on the spatiotemporal significant aggregation representation of the slurry flow rate, optimization instructions are obtained.

[0012] In the above-mentioned method for preparing the waste incineration flue gas denitrification catalyst, after the time queue of the slurry flow rate value is serially segmented, the local time queue of each of the slurry flow rate values ​​is serially encoded to obtain a time series of the local time series associated feature vector of the slurry flow rate, including: serially segmenting the time queue of the slurry flow rate value, and inputting the local time queue of each of the slurry flow rate values ​​into a sequence encoder based on the RNN model to obtain a time series of the local time series associated feature vector of the slurry flow rate.

[0013] In the above-mentioned method for preparing the waste incineration flue gas denitrification catalyst, the sequence of the historical slurry flow rate local time series associated feature vectors and the current slurry flow rate local time series associated feature vectors are subjected to feature significant aggregation based on energy transfer to obtain the spatiotemporal significant aggregation representation of the slurry flow rate, including: calculating the slurry flow rate static energy value of each historical slurry flow rate local time series associated feature vector in the sequence of the historical slurry flow rate local time series associated feature vectors; based on the slurry flow rate static energy value of each historical slurry flow rate local time series associated feature vector, the sequence of the historical slurry flow rate local time series associated feature vectors and the current slurry flow rate local time series associated feature vector are subjected to feature message transfer optimization to obtain the spatiotemporal significant aggregation representation of the slurry flow rate.

[0014] In the above-mentioned preparation method of the waste incineration flue gas denitrification catalyst, based on the slurry flow rate static energy value of each historical slurry flow rate local time series associated feature vector, the sequence of the historical slurry flow rate local time series associated feature vector and the current slurry flow rate local time series associated feature vector are subjected to feature message transmission optimization to obtain the slurry flow rate spatiotemporal significant aggregation representation, including: based on the slurry flow rate static energy value of each historical slurry flow rate local time series associated feature vector, the slurry flow rate dynamic energy spatiotemporal transfer value of each historical slurry flow rate local time series associated feature vector relative to the current slurry flow rate local time series associated feature vector is calculated; The slurry flow rate dynamic energy spatiotemporal transfer values ​​of each of the historical slurry flow rate local time series associated characteristic vectors are input into an energy transfer gating module based on a mask function to obtain a sequence of slurry flow rate dynamic energy spatiotemporal transfer weights; based on the sequence of slurry flow rate dynamic energy spatiotemporal transfer weights, the weighted sum of the sequence of historical slurry flow rate local time series associated characteristic vectors is calculated to obtain a historical slurry flow rate transfer aggregation modulation vector; the positional sum of the historical slurry flow rate transfer aggregation modulation vector and the current slurry flow rate local time series associated characteristic vector is calculated to obtain a slurry flow rate spatiotemporal significant aggregation representation vector as the slurry flow rate spatiotemporal significant aggregation representation.

[0015] In the above-mentioned preparation method of the waste incineration flue gas denitrification catalyst, based on the slurry flow rate static energy value of each historical slurry flow rate local time series correlation characteristic vector, the slurry flow rate dynamic energy spatiotemporal transfer value of each historical slurry flow rate local time series correlation characteristic vector relative to the current slurry flow rate local time series correlation characteristic vector is calculated, including: based on the spatial span coefficient between each historical slurry flow rate local time series correlation characteristic vector and the current slurry flow rate local time series correlation characteristic vector, determining the slurry flow rate energy spatial attenuation factor; based on each historical slurry flow rate local time series correlation characteristic vector, determining the slurry flow rate energy spatial attenuation factor; based on the spatial span coefficient between each historical slurry flow rate local time series correlation characteristic vector and the current slurry flow rate local time series correlation characteristic vector. The time span coefficient between the historical slurry flow velocity local time series correlation characteristic vector and the current slurry flow velocity local time series correlation characteristic vector is used to determine the slurry flow velocity energy time attenuation factor; based on the slurry flow velocity space attenuation factor and the slurry flow velocity energy time attenuation factor, the slurry flow velocity static energy value of each historical slurry flow velocity local time series correlation characteristic vector is temporally and spatially modulated to obtain the slurry flow velocity dynamic energy space-time transfer value of each historical slurry flow velocity local time series correlation characteristic vector relative to the current slurry flow velocity local time series correlation characteristic vector.

[0016] In the above-mentioned method for preparing the waste incineration flue gas denitrification catalyst, the slurry flow rate energy spatial attenuation factor is determined based on the spatial span coefficient between the local time series correlation characteristic vectors of each historical slurry flow rate and the local time series correlation characteristic vector of the current slurry flow rate, including: calculating the square of the spatial span value between the local time series correlation characteristic vector of the historical slurry flow rate and the local time series correlation characteristic vector of the current slurry flow rate to obtain the slurry flow rate spatial span coefficient; dividing the slurry flow rate static energy value corresponding to the local time series correlation characteristic vector of the historical slurry flow rate by the slurry flow rate spatial span coefficient to obtain the slurry flow rate energy spatial attenuation factor.

[0017] In the above-mentioned method for preparing the waste incineration flue gas denitrification catalyst, the slurry flow rate energy time decay factor is determined based on the time span coefficient between the local time series correlation characteristic vectors of each historical slurry flow rate and the local time series correlation characteristic vector of the current slurry flow rate, including: subtracting the slurry flow rate timestamp of the current slurry flow rate local time series correlation characteristic vector from the slurry flow rate timestamp of the historical slurry flow rate local time series correlation characteristic vector and rounding down to obtain the slurry flow rate time span value; dividing the slurry flow rate time span value by the time decay parameter to obtain the slurry flow rate time series span energy decay coefficient; using the slurry flow rate time series span energy decay coefficient as the exponential power, calculating the natural exponential function value with the natural constant e as the base to obtain the slurry flow rate time span coefficient; dividing the slurry flow rate static energy value corresponding to the historical slurry flow rate local time series correlation characteristic vector by the slurry flow rate time span coefficient to obtain the slurry flow rate energy time decay factor.

[0018] In the above-mentioned method for preparing the waste incineration flue gas denitrification catalyst, the slurry flow rate dynamic energy spatiotemporal transfer value of each historical slurry flow rate local time series associated characteristic vector is input into an energy transfer gating module based on a mask function to obtain a sequence of slurry flow rate dynamic energy spatiotemporal transfer weights, including: in response to the slurry flow rate dynamic energy spatiotemporal transfer value being greater than a predetermined threshold, the slurry flow rate dynamic energy spatiotemporal transfer value is input into a sigmoid function to obtain a value as the slurry flow rate dynamic energy spatiotemporal transfer weight.

[0019] In the above-mentioned method for preparing the waste incineration flue gas denitrification catalyst, an optimization instruction is obtained based on the spatiotemporal significant aggregation representation of the slurry flow rate, including: inputting the spatiotemporal significant aggregation representation vector of the slurry flow rate into a decoder-based magnetic field strength optimization control module to obtain the optimization instruction, and the optimization instruction includes a decoded value of the recommended magnetic field strength value at the next time point.

[0020] According to another aspect of the present application, a waste incineration flue gas denitration catalyst is provided, and the waste incineration flue gas denitration catalyst is prepared by the preparation method of the waste incineration flue gas denitration catalyst as described above.

[0021] This application has significant technical effects due to the adoption of the above technical solutions:

[0022] The waste incineration flue gas denitrification catalyst and its preparation method provided by the present application collect the time queue of the slurry flow rate value through the flow rate sensor, and use the data processing and analysis algorithm based on the deep neural network to segment and analyze the local time series association of the slurry flow rate value, so as to automatically obtain the recommended magnetic field strength value at the next time point according to the spatiotemporal significant aggregation between the local time series association characteristics of each slurry flow rate value. In this way, the magnetic field strength can be automatically adjusted in response to the change of the slurry flow rate, ensuring that the iron oxide separation process is always in the optimal state, reducing the need for manual intervention, and improving the automation level of the magnetic separation process. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0024] Figure 1 The present invention is a flow chart of a method for preparing a waste incineration flue gas denitration catalyst according to an embodiment of the present application.

[0025] Figure 2The present invention is a flow chart of a method for preparing a waste incineration flue gas denitration catalyst according to an embodiment of the present application, in which iron oxide is magnetically separated from the uniformly mixed red mud by a high-power magnetic separator.

[0026] Figure 3 This is a schematic diagram of data flow for magnetically separating the iron oxide in the uniformly mixed red mud by a high-power magnetic separator in the method for preparing a waste incineration flue gas denitrification catalyst according to an embodiment of the present application.

[0027] Figure 4 In the method for preparing a waste incineration flue gas denitrification catalyst according to an embodiment of the present application, the time series of the local time series correlation feature vector of the slurry flow rate is subjected to node spatiotemporal significant aggregation to obtain a flow chart of the spatiotemporal significant aggregation representation of the slurry flow rate.

[0028] Figure 5 In the method for preparing a waste incineration flue gas denitrification catalyst according to an embodiment of the present application, the sequence of the historical slurry flow rate local time series correlation feature vectors and the current slurry flow rate local time series correlation feature vectors are subjected to feature significant aggregation based on energy transfer to obtain a flow chart of the spatiotemporal significant aggregation representation of the slurry flow rate.

[0029] Figure 6 It is a schematic diagram of the process of treating waste incineration flue gas according to other embodiments of the present application. DETAILED DESCRIPTION

[0030] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0031] Waste incineration power generation is a process of converting waste into energy, but in this process, pollutants such as nitrogen oxides, VOCs gas and dioxins are produced, and the pollutants produced pose a threat to the environment and human health. In order to reduce the emission of these pollutants, patent CN114225936B proposes a waste incineration flue gas denitrification catalyst and its preparation method, which uses magnetic separation technology to extract α-Fe2O3 from red mud, and after a series of treatment steps, a mixed catalyst of α-Fe2O3 and γ-Fe2O3 is prepared to achieve low-temperature catalytic denitrification. In the magnetic separation process, precise control of the magnetic field strength is crucial for the effective recovery of iron oxide. However, the traditional magnetic separation method relies on manual adjustment, which is not only difficult to achieve high-precision control, but also prone to errors during frequent adjustments. In addition, the intelligence level of this method is not high, which increases the workload of operators and may cause misoperation due to fatigue.

[0032] Based on this, the present application provides a method for preparing a waste incineration flue gas denitrification catalyst. Figure 1 Flow chart of the preparation method of the waste incineration flue gas denitration catalyst according to the embodiment of the present application. Figure 1 As shown, the preparation method of the waste incineration flue gas denitrification catalyst according to the embodiment of the present application includes: S110, adding water to red mud and mixing evenly to obtain evenly mixed red mud; S120, magnetically selecting iron oxide in the evenly mixed red mud by a strong magnetic separator, and washing the screened iron oxide to obtain washed iron oxide; S130, ball milling and sieving the washed iron oxide to obtain sieved iron oxide; S140, calcining the sieved iron oxide in a reducing atmosphere to obtain the waste incineration flue gas denitrification catalyst.

[0033] It should be understood that red mud is an industrial solid waste generated by the alumina industry, which contains substances such as iron oxide and aluminum oxide. At present, the cumulative stockpile of red mud in the world has exceeded 4 billion tons, and the comprehensive utilization rate of red mud worldwide does not exceed 10%. While wasting resources, it also poses a potential serious threat to the ecological environment. In order to achieve the comprehensive utilization of red mud, it can be applied to the preparation of denitrification catalysts for waste incineration flue gas. In this application, it is first necessary to add water to the red mud and mix it evenly to obtain a uniformly mixed red mud. By adding water, the viscosity of the red mud is reduced, making it easier for the iron oxide particles to disperse in the slurry, thereby improving the effect of subsequent magnetic separation. Next, considering that most of the Fe2O3 in the red mud exists in the form of α-Fe2O3, which has weak magnetism, a strong magnetic separator can be used to separate Fe2O3 from the red mud, that is, the iron oxide in the mixed red mud is magnetically selected by a strong magnetic separator. Since the surface of the iron oxide may contain other soluble impurities, the screened iron oxide needs to be cleaned to obtain the cleaned iron oxide. Generally, the iron oxide is cleaned to a pH value of 7.0-7.5, at which time the impurity removal effect is better. Then, the cleaned iron oxide needs to be ball milled and sieved to obtain the sieved iron oxide. Here, by wet ball milling and sieving, the uniformity of heating of α-Fe2O3 during calcination can be improved, and then the uniformity of conversion of α-Fe2O3 to γ-Fe2O3 can be improved, so that when calcined at a specific temperature for a specific time, the mass ratio of α-Fe2O3 and γ-Fe2O3 can be controlled, thereby achieving low-temperature catalytic denitrification and dioxin removal. Finally, the sieved iron oxide is calcined in a reducing atmosphere to obtain the waste incineration flue gas denitration catalyst. Under specific reducing atmosphere conditions, the sieved iron oxide is calcined at high temperature to form a catalyst with specific structure and performance. Here, the reducing atmosphere specifically refers to a mixed gas of hydrogen and nitrogen.

[0034] The nitrogen oxides generated during the waste incineration process can achieve efficient denitration reaction, reduce denitration costs, improve denitration efficiency, and be more conducive to flue gas purification, reducing harm to the environment and human health under the catalytic action of the waste incineration flue gas denitration catalyst obtained through the preparation method.

[0035] Since the iron oxide (Fe2O3) in red mud mainly exists in the form of α-Fe2O3, this form of iron oxide has weak magnetism. Therefore, the strong magnetic separator can use its strong magnetic field to attract these weakly magnetic iron oxide particles, while other non-magnetic or weakly magnetic impurities will not be adsorbed, thereby effectively separating iron oxide from red mud. In the process of separating iron oxide from red mud using a strong magnetic separator, the magnetic field strength is a key parameter. The magnetic field strength directly affects the effect of magnetic separation, that is, the recovery rate and purity of iron oxide. Specifically, the appropriate magnetic field strength can ensure that as much iron oxide as possible is recovered from the red mud, thereby improving the efficiency of the entire process. If the magnetic field strength is too low, some iron oxide may not be attracted out, thereby reducing the recovery rate. Excessively high magnetic field strength may cause some non-magnetic or weakly magnetic impurities to be adsorbed, affecting the purity of the final product.

[0036] Therefore, in the process of magnetically separating the iron oxide in the mixed red mud by a strong magnetic separator, the technical concept of the present application is to collect the time queue of the slurry flow rate value by the flow rate sensor, and use the data processing and analysis algorithm based on the deep neural network to segment and analyze the local time series association of the slurry flow rate value, so as to automatically obtain the recommended magnetic field strength value at the next time point according to the spatiotemporal significant aggregation between the local time series association characteristics of each slurry flow rate value. In this way, the magnetic field strength can be automatically adjusted in response to the change of the slurry flow rate, ensuring that the separation process is always in the optimal state, reducing the need for manual intervention, and improving the automation level of the magnetic separation process.

[0037] Figure 2 The present invention is a flow chart of a method for preparing a waste incineration flue gas denitration catalyst according to an embodiment of the present application, in which iron oxide is magnetically separated from the uniformly mixed red mud by a high-power magnetic separator. Figure 3 The data flow diagram of the method for preparing a waste incineration flue gas denitration catalyst according to an embodiment of the present application is a diagram of the magnetic separation of iron oxide from the uniformly mixed red mud by a strong magnetic separator. Figure 2 and Figure 3As shown, the iron oxide in the uniformly mixed red mud is magnetically separated by a high-power magnetic separator, including: S121, obtaining a time queue of slurry flow rate values ​​collected by a flow rate sensor; S122, after sequence segmentation of the time queue of the slurry flow rate values, sequence encoding of each local time queue of the slurry flow rate values ​​to obtain a time series of local time series associated feature vectors of slurry flow rate; S123, performing node spatiotemporal significant aggregation on the time series of local time series associated feature vectors of slurry flow rate to obtain a spatiotemporal significant aggregation representation of slurry flow rate; S124, obtaining optimization instructions based on the spatiotemporal significant aggregation representation of slurry flow rate.

[0038] In step S121, the time queue of the slurry flow rate value collected by the flow rate sensor is obtained. It should be understood that the slurry flow rate directly affects the residence time and motion state of iron oxide in the strong magnetic separator. When the slurry flow rate increases, the speed of the iron oxide particles passing through the strong magnetic separator is accelerated. At this time, a stronger magnetic field strength is required to ensure that the iron oxide particles can be effectively adsorbed and separated in a shorter time; when the slurry flow rate decreases, the residence time of the iron oxide particles in the strong magnetic separator becomes longer. At this time, it is necessary to appropriately reduce the magnetic field strength, so that the iron oxide particles can be fully adsorbed and excessive energy consumption and equipment loss can be avoided. Considering that the slurry flow rate is a dynamically changing parameter, in order to ensure the stability of the magnetic separation effect under different slurry flow rate conditions, the present application is implemented by obtaining the flow rate value time series data and analyzing it to achieve magnetic field strength control. By real-time monitoring of the flow rate to help timely adjust the magnetic field strength to cope with the fluctuation of the flow rate, the intelligent level of magnetic separation operation can be improved, and the uncertainty caused by human factors can be reduced.

[0039] In step S122, after the time queue of the slurry flow rate value is segmented into a sequence, the local time queue of each of the slurry flow rate values ​​is sequence-encoded to obtain a time series of the local time series associated feature vector of the slurry flow rate. Specifically, in an embodiment of the present application, after the time queue of the slurry flow rate value is segmented into a sequence, the local time queue of each of the slurry flow rate values ​​is sequence-encoded to obtain a time series of the local time series associated feature vector of the slurry flow rate, including: segmenting the time queue of the slurry flow rate value into a sequence encoder based on the RNN model to obtain a time series of the local time series associated feature vector of the slurry flow rate. It should be understood that the local time series characteristics within each local time scale in the slurry flow rate value are different, such as the fluctuation and change trend of the short-term pattern. Based on this, in order to better capture the velocity characteristics and time dependence of each local time series, in the technical solution of the present application, the time queue of the slurry flow rate value is segmented into a sequence. Considering that the RNN model can capture the dynamic changes and long-term dependencies in the input time series data by hiding the state memory of the previous information, in order to capture and refine the timing dependencies and timing changes of the flow rate within each local time segment, the local time queue of each slurry flow rate value is further input into the sequence encoder based on the RNN model to obtain the time series of the local timing correlation feature vector of the slurry flow rate.

[0040] In step S123, the time series of the local time series associated feature vector of the slurry flow velocity is subjected to node spatiotemporal significant aggregation to obtain a spatiotemporal significant aggregation representation of the slurry flow velocity. Accordingly, considering that the changes of the slurry flow velocity at different time and space positions are complex, and the importance and influence of each time point in the entire time domain are different. Based on this, in order to effectively aggregate information from various time and space positions, so as to comprehensively learn the velocity pattern of the slurry flow velocity changing with time, to better integrate and characterize the aggregated information of the entire time domain, in the technical solution of the present application, the time series of the local time series associated feature vector of the slurry flow velocity is subjected to node spatiotemporal significant aggregation to obtain a slurry flow velocity spatiotemporal significant aggregation representation vector as a spatiotemporal significant aggregation representation of the slurry flow velocity.

[0041] Specifically, first, the sequence of the current slurry flow velocity local time series correlation feature vector and the historical slurry flow velocity local time series correlation feature vector is extracted from the time series of the slurry flow velocity local time series correlation feature vector as the sequence of the current node feature vector and the historical node feature vector, so that the current and future slurry flow velocity change trend can be better understood and predicted, the spatiotemporal dynamic relationship can be captured, and how the current situation is affected by history can be understood. Then, the static energy value of each historical node feature vector is calculated to obtain a sequence of slurry flow velocity static energy values. In particular, the slurry flow velocity static energy value here is obtained based on the maximum value, mean and variance of the historical slurry flow velocity local time series correlation feature vector. The obtained static energy value constructs a quantitative indicator to measure the essential attributes of the feature vector within the spatial range, establishes a reference point for the spatiotemporal decay process of potential energy, and helps the model grasp the relative importance of each node in the two dimensions of space and time. Next, based on the spatial span coefficient between the local time series correlation characteristic vectors of each historical slurry flow velocity and the local time series correlation characteristic vector of the current slurry flow velocity, the spatial attenuation factor of the slurry flow velocity energy is determined, and based on the time span coefficient between the local time series correlation characteristic vectors of each historical slurry flow velocity and the local time series correlation characteristic vector of the current slurry flow velocity, the time attenuation factor of the slurry flow velocity energy is determined. Specifically, the spatial span coefficient can evaluate the degree of influence of the historical slurry flow velocity characteristics on the current slurry flow velocity characteristics in space; the slurry flow velocity energy spatial attenuation factor can be used to determine the weight of historical data in the current data prediction, and the farther the historical data is from the current node, the smaller its influence. The time span coefficient can evaluate the degree of influence of the historical slurry flow velocity characteristics on the current slurry flow velocity characteristics in time; the slurry flow velocity energy time attenuation factor is used to determine the weight of historical data in the current data prediction, and the older the historical data is, the smaller its influence is. In this way, the sensitivity of the model to local and time factors is improved, making the transmission of information closer to real physical phenomena. Afterwards, based on the slurry flow rate energy space attenuation factor and the slurry flow rate energy time attenuation factor, the slurry flow rate static energy of each historical node feature vector is spatiotemporally modulated to obtain the slurry flow rate dynamic energy spatiotemporal transfer value, so as to combine the space and time factors and reflect the dynamic changes of the historical nodes in space and time. Then, the slurry flow rate dynamic energy spatiotemporal transfer value of each historical node feature vector is input into the energy transfer gating module based on the mask function to perform adaptive adjustment of the significant mask weight to obtain a sequence of slurry flow rate dynamic energy spatiotemporal transfer weights. Then, based on the sequence of weights, the sequence of historical node feature vectors is weighted fused to integrate information from different historical nodes to obtain the historical slurry flow rate transfer aggregation modulation vector. Finally, the historical modulation vector and the current slurry flow rate local time series correlation feature vector are summed by position to obtain the slurry flow rate spatiotemporal significant aggregation representation vector containing historical significant information and current information.

[0042] Specifically, Figure 4 The present invention is a flowchart of a method for preparing a waste incineration flue gas denitration catalyst according to an embodiment of the present application, in which the time series of the local time series correlation feature vector of the slurry flow rate is subjected to node spatiotemporal significant aggregation to obtain a spatiotemporal significant aggregation representation of the slurry flow rate. Figure 4 As shown, the time series of the local time series correlation feature vector of the slurry flow velocity is subjected to node spatiotemporal significant aggregation to obtain a spatiotemporal significant aggregation representation of the slurry flow velocity, including: S1231, extracting the current local time series correlation feature vector of the slurry flow velocity from the time series of the local time series correlation feature vector of the slurry flow velocity, and defining other local time series correlation feature vectors of the slurry flow velocity in the time series of the local time series correlation feature vector of the slurry flow velocity as historical local time series correlation feature vectors of the slurry flow velocity to obtain a sequence of historical local time series correlation feature vectors of the slurry flow velocity; S1232, performing feature significant aggregation based on energy transfer on the sequence of the historical local time series correlation feature vectors of the slurry flow velocity and the current local time series correlation feature vector of the slurry flow velocity to obtain the spatiotemporal significant aggregation representation of the slurry flow velocity.

[0043] More specifically, Figure 5 In the method for preparing a waste incineration flue gas denitration catalyst according to an embodiment of the present application, a flow chart is provided for performing energy transfer-based feature significant aggregation on the sequence of the local time series correlation feature vectors of the historical slurry flow rate and the local time series correlation feature vectors of the current slurry flow rate to obtain a time-space significant aggregation representation of the slurry flow rate. Figure 5 As shown, the sequence of the historical slurry flow velocity local time series associated feature vectors and the current slurry flow velocity local time series associated feature vector are subjected to feature significant aggregation based on energy transfer to obtain the spatiotemporal significant aggregation representation of the slurry flow velocity, including: S12321, calculating the slurry flow velocity static energy value of each historical slurry flow velocity local time series associated feature vector in the sequence of the historical slurry flow velocity local time series associated feature vectors; S12322, based on the slurry flow velocity static energy value of each historical slurry flow velocity local time series associated feature vector, performing feature message transfer optimization on the sequence of the historical slurry flow velocity local time series associated feature vectors and the current slurry flow velocity local time series associated feature vector to obtain the spatiotemporal significant aggregation representation of the slurry flow velocity.

[0044] More specifically, in the embodiment of the present application, the slurry velocity static energy value of each historical slurry velocity local time series associated characteristic vector in the sequence of the historical slurry velocity local time series associated characteristic vector is calculated, including: extracting the maximum value of the historical slurry velocity local time series associated characteristic vector to obtain the maximum value of the historical slurry velocity local time series characteristic; calculating the mean and variance of the historical slurry velocity local time series associated characteristic vector respectively to obtain the mean value of the historical slurry velocity local time series characteristic and the variance of the historical slurry velocity local time series characteristic; comparing the historical slurry velocity local time series associated characteristic vector with the historical slurry velocity local time series associated characteristic vector; The characteristic means are subtracted by position, and the fourth power of the characteristic value of each position in the local time series difference vector of the historical slurry flow velocity is calculated to obtain the local time series difference modulation vector of the historical slurry flow velocity; the expected value of the local time series difference modulation vector of the historical slurry flow velocity is calculated, and the expected value obtained is divided by the square of the variance of the local time series characteristic of the historical slurry flow velocity to obtain the local time series energy value of the historical slurry flow velocity; the maximum value of the local time series characteristic of the historical slurry flow velocity is added to the local time series energy value of the historical slurry flow velocity to obtain the static energy value of the slurry flow velocity corresponding to the local time series associated characteristic vector of the historical slurry flow velocity.

[0045] More specifically, in an embodiment of the present application, based on the slurry flow rate static energy values ​​of the respective historical slurry flow rate local time series associated characteristic vectors, the sequence of the historical slurry flow rate local time series associated characteristic vectors and the current slurry flow rate local time series associated characteristic vector are subjected to feature message transmission optimization to obtain the spatiotemporal significant aggregation representation of the slurry flow rate, including: based on the slurry flow rate static energy values ​​of the respective historical slurry flow rate local time series associated characteristic vectors, calculating the spatiotemporal dynamic energy transfer value of the slurry flow rate of the respective historical slurry flow rate local time series associated characteristic vectors relative to the current slurry flow rate local time series associated characteristic vector; The slurry flow rate dynamic energy spatiotemporal transfer values ​​of each historical slurry flow rate local time series associated characteristic vector are input into the energy transfer gating module based on the mask function to obtain a sequence of slurry flow rate dynamic energy spatiotemporal transfer weights; based on the sequence of slurry flow rate dynamic energy spatiotemporal transfer weights, the weighted sum of the sequence of historical slurry flow rate local time series associated characteristic vectors is calculated to obtain a historical slurry flow rate transfer aggregation modulation vector; the positional sum of the historical slurry flow rate transfer aggregation modulation vector and the current slurry flow rate local time series associated characteristic vector is calculated to obtain a slurry flow rate spatiotemporal significant aggregation representation vector as the slurry flow rate spatiotemporal significant aggregation representation.

[0046] More specifically, in the embodiment of the present application, based on the static energy value of the slurry flow rate of each historical slurry flow rate local time series associated characteristic vector, the slurry flow rate dynamic energy spatiotemporal transfer value of each historical slurry flow rate local time series associated characteristic vector relative to the current slurry flow rate local time series associated characteristic vector is calculated, including: based on the spatial span coefficient between the each historical slurry flow rate local time series associated characteristic vector and the current slurry flow rate local time series associated characteristic vector, determining the slurry flow rate energy spatial attenuation factor; based on the each historical slurry flow rate local time series associated characteristic vector The slurry flow velocity local time series correlation characteristic vector and the current slurry flow velocity local time series correlation characteristic vector are used to determine the slurry flow velocity energy time attenuation factor; based on the slurry flow velocity energy spatial attenuation factor and the slurry flow velocity energy time attenuation factor, the slurry flow velocity static energy value of each historical slurry flow velocity local time series correlation characteristic vector is temporally and spatially modulated to obtain the slurry flow velocity dynamic energy space-time transfer value of each historical slurry flow velocity local time series correlation characteristic vector relative to the current slurry flow velocity local time series correlation characteristic vector.

[0047] More specifically, in an embodiment of the present application, based on the spatial span coefficient between the each historical slurry flow velocity local time series correlation characteristic vector and the current slurry flow velocity local time series correlation characteristic vector, the slurry flow velocity energy spatial attenuation factor is determined, including: calculating the square of the spatial span value between the historical slurry flow velocity local time series correlation characteristic vector and the current slurry flow velocity local time series correlation characteristic vector to obtain the slurry flow velocity spatial span coefficient; dividing the slurry flow velocity static energy value corresponding to the historical slurry flow velocity local time series correlation characteristic vector by the slurry flow velocity spatial span coefficient to obtain the slurry flow velocity energy spatial attenuation factor.

[0048] More specifically, in an embodiment of the present application, based on the time span coefficient between the each historical slurry flow rate local time series associated characteristic vector and the current slurry flow rate local time series associated characteristic vector, the slurry flow rate energy time decay factor is determined, including: subtracting the slurry flow rate timestamp of the current slurry flow rate local time series associated characteristic vector from the slurry flow rate timestamp of the historical slurry flow rate local time series associated characteristic vector, and rounding down to obtain a slurry flow rate time span value; dividing the slurry flow rate time span value by a time decay parameter to obtain a slurry flow rate time series span energy decay coefficient; using the slurry flow rate time series span energy decay coefficient as an exponential power, calculating a natural exponential function value with a natural constant e as the base to obtain a slurry flow rate time span coefficient; dividing the slurry flow rate static energy value corresponding to the historical slurry flow rate local time series associated characteristic vector by the slurry flow rate time span coefficient to obtain the slurry flow rate energy time decay factor.

[0049] More specifically, in an embodiment of the present application, based on the slurry flow velocity energy spatial attenuation factor and the slurry flow velocity energy temporal attenuation factor, the slurry flow velocity static energy value of each historical slurry flow velocity local time series associated characteristic vector is temporally modulated to obtain the slurry flow velocity dynamic energy spatiotemporal transfer value of each historical slurry flow velocity local time series associated characteristic vector relative to the current slurry flow velocity local time series associated characteristic vector, including: multiplying the slurry flow velocity energy spatial attenuation factor and the slurry flow velocity energy temporal attenuation factor of the historical slurry flow velocity local time series associated characteristic vector with the first modulation parameter and the second modulation parameter respectively, and adding the obtained slurry flow velocity energy spatial attenuation modulation factor and slurry flow velocity energy temporal attenuation modulation factor to obtain the slurry flow velocity dynamic energy spatiotemporal transfer value.

[0050] More specifically, in an embodiment of the present application, the slurry flow rate dynamic energy spatiotemporal transfer values ​​of the respective historical slurry flow rate local time series associated characteristic vectors are input into an energy transfer gating module based on a mask function to obtain a sequence of slurry flow rate dynamic energy spatiotemporal transfer weights, including: in response to the slurry flow rate dynamic energy spatiotemporal transfer value being greater than a predetermined threshold, the slurry flow rate dynamic energy spatiotemporal transfer value is input into a sigmoid function to obtain a value as the slurry flow rate dynamic energy spatiotemporal transfer weight.

[0051] In the embodiment of the present application, specifically, the time series of the local time series correlation feature vector of the slurry flow rate is subjected to node spatiotemporal significant aggregation, which can be expressed by the formula:

[0052] X={x1,x2,…,x i , …, xp}

[0053]

[0054] Wherein, X is the time series of the local time series correlation feature vector of the slurry flow rate, x1, x2, ..., xi, ..., xp are the 1st, 2nd, ..., ith, ..., pth local time series correlation feature vectors of the slurry flow rate in the time series of the local time series correlation feature vector of the slurry flow rate, respectively, xi(j) is the eigenvalue of each position in the ith local time series correlation feature vector of the slurry flow rate, E(·) is the calculated expected value, μi and σi2 are the mean and variance of xi, respectively, and max(xi) is the maximum value extracted from xi. is the static energy value of the slurry flow velocity corresponding to xi, Count(xi→xp) represents the spatial span value between xi and xp, is the spatial attenuation factor of the slurry velocity energy between xi and xp, tp and ti distributions represent the timestamps of the pth and ith slurry velocity local time series correlation feature vectors, is a floor operation, L is the time decay parameter, exp(·) represents the exponential function value with the natural constant e as the base, is the time attenuation factor of the slurry velocity energy between xi and xp, γ and δ are the corresponding modulation parameters, e(i→p) is the spatiotemporal transfer value of the slurry velocity dynamic energy between xi and xp, mas k (·) is masking processing, sigmoid (·) is the sigmoid function, θ is a predetermined threshold, p is the number of vectors in the time series of the local time series correlation feature vector of the slurry flow rate, x c is the spatiotemporal significant aggregation vector of the slurry flow rate.

[0055] In step S124, based on the spatiotemporal significant aggregation representation of the slurry flow rate, an optimization instruction is obtained. Specifically, in an embodiment of the present application, based on the spatiotemporal significant aggregation representation of the slurry flow rate, an optimization instruction is obtained, including: inputting the spatiotemporal significant aggregation representation vector of the slurry flow rate into a decoder-based magnetic field strength optimization control module to obtain the optimization instruction, and the optimization instruction includes a decoded value of the recommended magnetic field strength value at the next time point. That is, the spatiotemporal significant aggregation representation of the slurry flow rate obtained by performing spatiotemporal significant aggregation of nodes using the time series of the local temporal association feature vector of the slurry flow rate is decoded to automatically obtain the recommended magnetic field strength value at the next time point. In this way, the magnetic field strength can be automatically adjusted in response to changes in the slurry flow rate, ensuring that the separation process is always in the optimal state, reducing the need for manual intervention, and improving the automation level of the magnetic separation process.

[0056] In particular, in the present application, taking into account the time series of the local temporal correlation feature vector of the slurry flow rate, the temporal correlation characteristics of the slurry flow rate value in the local time domain are represented, and when the nodes are significantly aggregated in time and space, based on the time and space transmission differences of the temporal node potential energy between the local time domains, the slurry flow rate time and space significant aggregation representation vector will also have a global time domain propagation-type aggregation difference. Therefore, it is expected to improve the detailed temporal semantic aggregation expression effect of the slurry flow rate time and space significant aggregation representation vector based on the local time domain temporal feature propagation-type aggregation difference under the global time domain.

[0057] Preferably, the slurry flow rate spatiotemporal significant aggregation representation vector is input into the decoder-based magnetic field strength optimization control module to obtain the optimization instruction, including: calculating the sum of the absolute values ​​of each eigenvalue of the slurry flow rate spatiotemporal significant aggregation representation vector to obtain a first slurry flow rate spatiotemporal significant aggregation and modulation value, and calculating the square root of the sum of the squares of each eigenvalue of the slurry flow rate spatiotemporal significant aggregation representation vector to obtain a second slurry flow rate spatiotemporal significant aggregation and modulation value; performing dot-wise subtraction of the slurry flow rate spatiotemporal significant aggregation representation vector and the second slurry flow rate spatiotemporal significant aggregation and modulation value, respectively performing dot-wise multiplication with the number of eigenvalues ​​of the slurry flow rate spatiotemporal significant aggregation representation vector and the reciprocal of the first slurry flow rate spatiotemporal significant aggregation and modulation value, and taking the reciprocal of each eigenvalue to obtain the first slurry flow rate spatiotemporal significant aggregation and modulation value. Empty significant aggregation phase transition vector; after point subtraction of the slurry flow velocity spatiotemporal significant aggregation representation vector and the first slurry flow velocity spatiotemporal significant aggregation and modulation value, respectively perform point multiplication with the square root of the number of eigenvalues ​​of the slurry flow velocity spatiotemporal significant aggregation representation vector and the inverse of the second slurry flow velocity spatiotemporal significant aggregation and modulation value, and take the inverse of each eigenvalue to obtain the second slurry flow velocity spatiotemporal significant aggregation phase transition vector; subtract the point product vector of the second slurry flow velocity spatiotemporal significant aggregation phase transition vector and the weighted hyperparameter from the first slurry flow velocity spatiotemporal significant aggregation phase transition vector to obtain an optimized slurry flow velocity spatiotemporal significant aggregation representation vector; input the optimized slurry flow velocity spatiotemporal significant aggregation representation vector into the decoder-based magnetic field strength optimization control module to obtain the optimization instruction.

[0058] Here, the slurry flow rate spatiotemporal significant aggregation representation vector, denoted as V, is optimally represented as:

[0059]

[0060] v i ∈V∈R n

[0061] Wherein, V represents the spatiotemporal significant aggregation representation vector of the slurry flow rate, vi is the eigenvalue of the ith position in the spatiotemporal significant aggregation representation vector of the slurry flow rate, n represents the number of eigenvalues ​​in the spatiotemporal significant aggregation representation vector of the slurry flow rate, α represents the spatiotemporal significant aggregation and modulation value of the first slurry flow rate, β represents the spatiotemporal significant aggregation and modulation value of the second slurry flow rate, ⊙ represents the point multiplication by position, Indicates point-by-point subtraction, [·] ⊙ - 1 is the inverse of each eigenvalue of the vector, V1 represents the first slurry flow rate spatiotemporal significant aggregation phase transition vector, V2 represents the second slurry flow rate spatiotemporal significant aggregation phase transition vector, ω represents a weighted hyperparameter, and V′ represents the optimized slurry flow rate spatiotemporal significant aggregation representation vector.

[0062] That is, the present application uses the difference of the characteristic value of the slurry flow rate spatiotemporal significant aggregation representation vector relative to the vector feature set of the slurry flow rate spatiotemporal significant aggregation representation vector as the semantic change intensity information, and performs a phase-like conversion corresponding to the position-based intensity modulation through different and modulation representation forms, so that the aggregation enhancement of the semantic change phase perception can improve the axial aggregation receptive field along the feature aggregation direction by performing a spatial translation operation based on alternating stacking under the vector set scale balance of the slurry flow rate spatiotemporal significant aggregation representation vector, thereby improving the perception effect of the aggregation semantics of the slurry flow rate spatiotemporal significant aggregation representation vector on the change of the detail time sequence semantics, so as to improve the expression effect of the slurry flow rate spatiotemporal significant aggregation representation vector, and improve the accuracy of the optimization instruction obtained by inputting the decoder-based magnetic field strength optimization control module. In this way, it is possible to quickly respond to the change of the slurry flow rate to automatically adjust the magnetic field strength, ensure that the separation process is always in the optimal state, reduce the need for manual intervention, and improve the automation level of the magnetic separation process.

[0063] In summary, the preparation method of the waste incineration flue gas denitrification catalyst based on the embodiment of the present application is explained, which collects the time queue of the slurry flow rate value through the flow rate sensor, and uses the data processing and analysis algorithm based on the deep neural network to segment and analyze the local time series association of the slurry flow rate value, so as to automatically obtain the recommended magnetic field strength value at the next time point according to the spatiotemporal significant aggregation between the local time series association characteristics of each slurry flow rate value. In this way, the magnetic field strength can be automatically adjusted in response to the change of the slurry flow rate, ensuring that the iron oxide separation process is always in the optimal state, reducing the need for manual intervention, and improving the automation level of the magnetic separation process.

[0064] In an embodiment of the present application, a waste incineration flue gas denitration catalyst is also provided, and the waste incineration flue gas denitration catalyst is prepared by the waste incineration flue gas denitration catalyst preparation method described above.

[0065] In other embodiments of the present application, a waste incineration flue gas treatment process is provided, which realizes flue gas purification through selective catalytic reduction denitrification (SCR) + semi-dry deacidification + activated carbon injection to remove dioxins and heavy metals + bag dust removal. It is particularly noteworthy that selective catalytic denitrification is used here, which has a higher denitrification efficiency than the traditional selective non-catalytic reduction denitrification (SNCR). Generally, a selective catalytic reduction denitrification device is added at the end of the process flow. The flue gas treatment process provided in this embodiment reasonably modifies the active sites of the medium-temperature denitrification vanadium titanium catalyst so that it has a high catalytic activity for VOCs at a reaction temperature of 200-240°C without reducing the denitrification activity and selectivity of the catalyst. At the same time, it ensures that the catalyst maintains a long-term catalytic activity in an atmosphere containing high water vapor, acidic gases, alkali metals and heavy metals, and realizes the regeneration of the deactivated catalyst after long-term operation. In the actual industrial application of waste incineration flue gas treatment technology, the process and equipment optimization of tail gas front-end deacidification, activated carbon injection and bag dust removal have been realized to minimize the content of water vapor, acid gas, alkali metal and heavy metal entering the synergistic catalytic device, extend the service life of the catalyst, and achieve the modular effect of synergistic catalysis through industrial application to achieve replicable production. Specifically, using existing process equipment, VOCs and NOx pollutants can be catalytically degraded into harmless or easily adsorbed compounds at the same time, thereby effectively reducing investment, operation and maintenance costs and saving site space.

[0066] Figure 6 FIG. 1 is a schematic diagram of a waste incineration flue gas treatment process according to another embodiment of the present application. Figure 6 As shown, the waste incineration flue gas treatment process includes: a. Waste incineration process: first, the garbage truck transports the garbage to the garbage unloading hall and dumps the garbage into the garbage pool, and then the garbage in the garbage pool is delivered to the grate furnace by the garbage crane, and the garbage is incinerated in the grate furnace, and the slag generated by the garbage incineration is transported out by truck; b. Steam process: The preheated gas generated by the air preheater treated by the gas inhaled by the primary blower and the gas directly inhaled by the secondary blower are directly gathered in the grate furnace, and the heat generated by the garbage incineration is used to heat Hot steam, the steam obtained by heating is used to drive the turbine to generate electricity; c. Flue gas treatment process: Some acidic gases produced by garbage incineration first flow into the deacidification reaction tower for deacidification treatment. The gas after deacidification can effectively remove dioxins and heavy metals through the adsorption of activated carbon and dry lime. At this time, ash will be formed and will be filtered out in the bag filter. The filtered ash will be transported out by truck through the ash bin. The gas obtained from layer by layer treatment is finally treated by SCR (selective catalytic reduction technology) to remove nitrogen oxides in the gas, and finally discharged through the chimney.

[0067] The basic principles of the present application are described above in conjunction with specific embodiments. However, it should be noted that the advantages, strengths, effects, etc. mentioned in the present application are only examples and not limitations, and it cannot be considered that these advantages, strengths, effects, etc. are required by each embodiment of the present application. In addition, the specific details of the above application are only for the purpose of illustration and ease of understanding, rather than limitation, and the above details do not limit the present application to the use of the above specific details to be implemented.

Claims

1. A method for preparing a waste incineration flue gas denitration catalyst, comprising: adding water to the red mud and mixing well to obtain well-mixed red mud; The iron oxide in the uniformly mixed red mud is magnetically separated by a strong magnetic separator, and the screened iron oxide is cleaned to obtain the cleaned iron oxide; the cleaned iron oxide is ball-milled and sieved to obtain the sieved iron oxide; the sieved iron oxide is calcined in a reducing atmosphere to obtain the waste incineration flue gas denitration catalyst; characterized in that the iron oxide in the uniformly mixed red mud is magnetically separated by a strong magnetic separator, comprising: Obtaining a time queue of slurry flow rate values ​​collected by a flow rate sensor; After the time queue of the slurry flow rate value is segmented into a sequence, the local time queue of each slurry flow rate value is sequence-encoded to obtain a time series of a local time series correlation feature vector of the slurry flow rate; The time series of the local time series correlation feature vectors of the slurry flow velocity is subjected to node spatiotemporal significant aggregation to obtain a spatiotemporal significant aggregation representation of the slurry flow velocity, including: extracting the current local time series correlation feature vector of the slurry flow velocity from the time series of the local time series correlation feature vectors of the slurry flow velocity, and defining other local time series correlation feature vectors of the slurry flow velocity in the time series of the local time series correlation feature vectors of the slurry flow velocity as historical local time series correlation feature vectors of the slurry flow velocity to obtain a sequence of historical local time series correlation feature vectors of the slurry flow velocity; and performing feature significant aggregation based on energy transfer on the sequence of the historical local time series correlation feature vectors of the slurry flow velocity and the current local time series correlation feature vector of the slurry flow velocity to obtain the spatiotemporal significant aggregation representation of the slurry flow velocity; Based on the spatiotemporal significant aggregation representation of the slurry flow rate, optimization instructions are obtained.

2. The method for preparing a waste incineration flue gas denitration catalyst according to claim 1, characterized in that: After the time queue of the slurry flow rate values ​​is segmented into a sequence, the local time queues of each of the slurry flow rate values ​​are sequence encoded to obtain a time series of the local time series associated feature vectors of the slurry flow rate, including: segmenting the time queue of the slurry flow rate values ​​into a sequence encoder based on the RNN model to obtain a time series of the local time series associated feature vectors of the slurry flow rate.

3. The method for preparing a waste incineration flue gas denitration catalyst according to claim 2, characterized in that: Performing energy transfer-based feature significant aggregation on the sequence of the historical slurry flow velocity local time series associated feature vectors and the current slurry flow velocity local time series associated feature vector to obtain the slurry flow velocity spatiotemporal significant aggregation representation, including: Calculating the slurry flow rate static energy value of each historical slurry flow rate local time series correlation feature vector in the sequence of the historical slurry flow rate local time series correlation feature vector; Based on the slurry flow rate static energy value of each historical slurry flow rate local time series associated feature vector, the sequence of the historical slurry flow rate local time series associated feature vector and the current slurry flow rate local time series associated feature vector are optimized by feature message passing to obtain the spatiotemporal significant aggregation representation of the slurry flow rate.

4. The method for preparing a waste incineration flue gas denitration catalyst according to claim 3, characterized in that: Based on the slurry flow rate static energy value of each historical slurry flow rate local time series associated feature vector, the sequence of the historical slurry flow rate local time series associated feature vector and the current slurry flow rate local time series associated feature vector are subjected to feature message transmission optimization to obtain the slurry flow rate spatiotemporal significant aggregation representation, including: Based on the slurry flow velocity static energy value of each historical slurry flow velocity local time series associated characteristic vector, calculate the slurry flow velocity dynamic energy spatiotemporal transfer value of each historical slurry flow velocity local time series associated characteristic vector relative to the current slurry flow velocity local time series associated characteristic vector; Inputting the slurry flow rate dynamic energy spatiotemporal transfer values ​​of the local time series associated characteristic vectors of each historical slurry flow rate into an energy transfer gating module based on a mask function to obtain a sequence of slurry flow rate dynamic energy spatiotemporal transfer weights; Based on the sequence of spatiotemporal transfer weights of the slurry flow rate dynamic energy, a weighted sum of the sequence of local time series associated characteristic vectors of the historical slurry flow rate is calculated to obtain a historical slurry flow rate transfer aggregation modulation vector; The position-wise sum of the historical slurry flow rate transfer aggregation modulation vector and the current slurry flow rate local time series correlation feature vector is calculated to obtain a slurry flow rate spatiotemporal significant aggregation representation vector as the slurry flow rate spatiotemporal significant aggregation representation.

5. The method for preparing a waste incineration flue gas denitration catalyst according to claim 4, characterized in that: Based on the slurry flow velocity static energy value of each historical slurry flow velocity local time series associated characteristic vector, the slurry flow velocity dynamic energy spatiotemporal transfer value of each historical slurry flow velocity local time series associated characteristic vector relative to the current slurry flow velocity local time series associated characteristic vector is calculated, including: Determining a slurry flow rate energy spatial attenuation factor based on a spatial span coefficient between each of the historical slurry flow rate local time series correlation feature vectors and the current slurry flow rate local time series correlation feature vector; Determining a slurry flow rate energy time attenuation factor based on a time span coefficient between each of the historical slurry flow rate local time series associated characteristic vectors and the current slurry flow rate local time series associated characteristic vector; Based on the slurry flow velocity energy spatial attenuation factor and the slurry flow velocity energy temporal attenuation factor, the slurry flow velocity static energy value of each historical slurry flow velocity local time series correlation characteristic vector is temporally and spatially modulated to obtain the slurry flow velocity dynamic energy spatiotemporal transfer value of each historical slurry flow velocity local time series correlation characteristic vector relative to the current slurry flow velocity local time series correlation characteristic vector.

6. The method for preparing a waste incineration flue gas denitration catalyst according to claim 5, characterized in that: Based on the spatial span coefficient between the local time series correlation characteristic vectors of each historical slurry flow rate and the local time series correlation characteristic vector of the current slurry flow rate, the slurry flow rate energy spatial attenuation factor is determined, including: Calculating the square of the spatial span value between the local time series correlation feature vector of the historical slurry flow rate and the local time series correlation feature vector of the current slurry flow rate to obtain the slurry flow rate spatial span coefficient; The slurry flow rate static energy value corresponding to the local time series correlation characteristic vector of the historical slurry flow rate is divided by the slurry flow rate spatial span coefficient to obtain the slurry flow rate energy spatial attenuation factor.

7. The method for preparing a waste incineration flue gas denitration catalyst according to claim 6, characterized in that: Determining the slurry flow rate energy time attenuation factor based on the time span coefficient between each of the historical slurry flow rate local time series correlation feature vectors and the current slurry flow rate local time series correlation feature vector includes: The slurry flow rate timestamp of the current slurry flow rate local time series associated feature vector is subtracted from the slurry flow rate timestamp of the historical slurry flow rate local time series associated feature vector, and then rounded down to obtain a slurry flow rate time span value; Dividing the slurry flow rate time span value by the time decay parameter to obtain the slurry flow rate time series span energy decay coefficient; Taking the energy attenuation coefficient of the slurry flow rate time series span as the exponential power, calculating the value of the natural exponential function with the natural constant e as the base to obtain the slurry flow rate time span coefficient; The slurry flow rate static energy value corresponding to the historical slurry flow rate local time series correlation characteristic vector is divided by the slurry flow rate time span coefficient to obtain the slurry flow rate energy time attenuation factor.

8. The method for preparing a waste incineration flue gas denitration catalyst according to claim 7, characterized in that: The slurry flow rate dynamic energy spatiotemporal transfer values ​​of the respective historical slurry flow rate local time series associated characteristic vectors are input into an energy transfer gating module based on a mask function to obtain a sequence of slurry flow rate dynamic energy spatiotemporal transfer weights, including: in response to the slurry flow rate dynamic energy spatiotemporal transfer value being greater than a predetermined threshold, the slurry flow rate dynamic energy spatiotemporal transfer value is input into a sigmoid function to obtain a value as the slurry flow rate dynamic energy spatiotemporal transfer weight.

9. The method for preparing a waste incineration flue gas denitration catalyst according to claim 8, characterized in that: Based on the spatiotemporal significant aggregation representation of the slurry flow rate, an optimization instruction is obtained, including: inputting the spatiotemporal significant aggregation representation vector of the slurry flow rate into a decoder-based magnetic field strength optimization control module to obtain the optimization instruction, wherein the optimization instruction includes a decoded value of a recommended magnetic field strength value at the next time point.

10. A waste incineration flue gas denitrification catalyst, characterized in that: The waste incineration flue gas denitration catalyst is prepared by the preparation method of the waste incineration flue gas denitration catalyst according to any one of claims 1-9.

Images