An industrial pollution source monitoring method and system

Abstract

An industrial pollution source monitoring method and system provided by the present invention relate to the technical field of data processing. In the present invention, a target pollution data set and a historical pollution data set are compared and analyzed to output first pollution change amplitude information. According to the target pollution data set, data adjustment is performed to output an adjusted target pollution data set that matches the target pollution data set, and then according to the historical pollution data set, data adjustment is performed to output an adjusted historical pollution data set that matches the historical pollution data set. The adjusted target pollution data set and the adjusted historical pollution data set are compared and analyzed to output second pollution change amplitude information, and then the second pollution change amplitude information and the first pollution change amplitude information are fused to form target pollution change amplitude information. Based on the above method, the reliability of pollution change amplitude monitoring can be improved to a certain extent.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular, to an industrial pollution source monitoring method and system. Background Art

[0002] On the basis of the continuous development of industry, the problem of environmental pollution has become increasingly serious and has become a key concern of the general public. Among them, pollution sources are the sources of environmental pollution. Generally, pollution sources can include air pollution sources, water pollution sources, etc. Correspondingly, the content of environmental monitoring includes air pollution source monitoring, water pollution source monitoring, etc. For example, environmental monitoring equipment can be used to determine the emission sources, emission amounts, and types of pollutants of pollutants, providing strong data support for environmental monitoring and also providing a basis for controlling pollution source emissions and environmental pollution disputes.

[0003] In addition, in some applications, it may be necessary to monitor changes in the degree of pollution, such as comparing and analyzing pollution data for two adjacent times to determine the change range of the degree of pollution. However, only comparing and analyzing pollution data for two adjacent times may have the problem of low reliability. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to provide an industrial pollution source monitoring method and system to improve the reliability of monitoring the change range of pollution to a certain extent.

[0005] To achieve the above purpose, the embodiments of the present invention adopt the following technical solutions:

[0006] An industrial pollution source monitoring method is applied to a pollution source monitoring server, and the pollution source monitoring server is communicatively connected to a plurality of industrial pollution source monitoring devices. The plurality of industrial pollution source monitoring devices are arranged in sequence to respectively collect pollution data for a corresponding industrial pollution area in a plurality of industrial pollution areas included in the industrial pollution source. The method includes:

[0007] Comparing and analyzing a target pollution data set and a historical pollution data set to output first pollution change range information. The target pollution data set includes multiple pieces of industrial pollution data collected by each industrial pollution source monitoring device among the plurality of industrial pollution source monitoring devices during the current time period, and the historical pollution data set includes multiple pieces of historical industrial pollution data collected by each industrial pollution source monitoring device among the plurality of industrial pollution source monitoring devices during the previous time period;

[0008] Based on the target pollution data set, data adjustment is performed to output an adjusted target pollution data set that matches the target pollution data set, and then based on the historical pollution data set, data adjustment is performed to output an adjusted historical pollution data set that matches the historical pollution data set;

[0009] Compare and analyze the adjusted target pollution data set and the adjusted historical pollution data set to output second pollution change range information, and then fuse the second pollution change range information and the first pollution change range information to form target pollution change range information.

[0010] In some preferred embodiments, in the above industrial pollution source monitoring method, the step of comparing and analyzing the target pollution data set and the historical pollution data set to output first pollution change range information includes:

[0011] According to the corresponding industrial pollution source monitoring equipment and the internal moments within the corresponding time period, perform one-to-one correspondence processing of the pollution data between the target pollution data set and the historical pollution data set, so that each industrial pollution data corresponds to a historical industrial pollution data, so that there are the same industrial pollution source monitoring equipment and internal moments within the time period between each industrial pollution data and the corresponding historical industrial pollution data, and the internal moments within the time period are used to reflect the sequence of the collection time of the corresponding pollution data within the corresponding time period;

[0012] Calculate the data difference values between each group of corresponding industrial pollution data and historical industrial pollution data respectively, and then fuse the data difference values between each group of corresponding industrial pollution data and historical industrial pollution data to output first pollution change range information.

[0013] In some preferred embodiments, in the above industrial pollution source monitoring method, the step of outputting an adjusted target pollution data set that matches the target pollution data set by performing data adjustment according to the target pollution data set includes:

[0014] According to the target pollution data set, form an ordered set of pollution data. The ordered set of pollution data includes multiple pollution data sorting queues corresponding to multiple internal time periods of the time period. Each internal time period of the time period includes multiple internal moments of the time period. At each internal moment of the time period, each industrial pollution source monitoring equipment collects pollution data for the corresponding industrial pollution area to form corresponding industrial pollution data. Each pollution data sorting queue includes a first number of rows and a second number of columns. The first number is equal to the number of internal moments of the time period, and the second number is equal to the number of industrial pollution source monitoring equipment. The multiple pollution data sorting queues include a first pollution data sorting queue, and the first pollution data sorting queue is any one of the multiple pollution data sorting queues;

[0015] From each of the pollution data sorting queues, identify multiple initial pollution data feature distributions with different characteristic dimension quantities, and then splice the feature distributions of the multiple initial pollution data feature distributions corresponding to one pollution data sorting queue to output the spliced pollution data feature distribution corresponding to each pollution data sorting queue;

[0016] According to the feature matching degree between the spliced pollution data feature distribution of each pollution data sorting queue and the spliced pollution data feature distribution of the first pollution data sorting queue, perform feature distribution splicing with splicing weights on the spliced pollution data feature distributions of the multiple pollution data sorting queues, and output the initial spliced pollution data feature distribution corresponding to the first pollution data sorting queue;

[0017] According to the initial spliced pollution data feature distribution, perform reduction processing on the sorting queue of the first pollution data sorting queue to output the reduced first pollution data sorting queue;

[0018] After successively taking each pollution data sorting queue in the multiple pollution data sorting queues as the first pollution data sorting queue for processing to output the reduced first pollution data sorting queue corresponding to each pollution data sorting queue, then combine them according to the reduced first pollution data sorting queue corresponding to each pollution data sorting queue to form an adjusted target pollution data set matching the target pollution data set.

[0019] In some preferred embodiments, in the above industrial pollution source monitoring method, the step of splicing the feature distributions of the multiple initial pollution data feature distributions corresponding to one pollution data sorting queue to output the spliced pollution data feature distribution corresponding to each pollution data sorting queue includes:

[0020] For any one pollution data sorting queue, according to the characteristic dimension quantity with the maximum value among the multiple characteristic dimension quantities, adjust the characteristic dimension quantity of each other initial pollution data feature distribution other than the initial pollution data feature distribution corresponding to the characteristic dimension quantity with the maximum value in the multiple initial pollution data feature distributions corresponding to this pollution data sorting queue to the characteristic dimension quantity with the maximum value, so as to form multiple adjusted initial pollution data feature distributions, such that the characteristic dimension quantity of each adjusted initial pollution data feature distribution is equal to the characteristic dimension quantity with the maximum value;

[0021] Fuse the multiple adjusted initial pollution data feature distributions to output the fused pollution data feature distribution corresponding to the multiple adjusted initial pollution data feature distributions;

[0022] Perform feature mining on the feature distribution of the fused pollution data, and output the spliced pollution data feature distribution corresponding to the pollution data sorting queue.

[0023] In some preferred embodiments, in the above industrial pollution source monitoring method, the step of respectively identifying multiple initial pollution data feature distributions with different feature dimension quantities from each of the pollution data sorting queues includes:

[0024] For each pollution data sorting queue among the multiple pollution data sorting queues, perform data feature mining on this pollution data sorting queue, and output the first pollution data feature distribution corresponding to this pollution data sorting queue;

[0025] Perform feature compression on the first pollution data feature distribution, and output the first compressed pollution data feature distribution with the first feature dimension quantity corresponding to the first pollution data feature distribution;

[0026] Perform a reduction process on the feature dimension quantity of the first compressed pollution data feature distribution with the first feature dimension quantity, and output the first compressed pollution data feature distribution with the second feature dimension quantity corresponding to the first pollution data feature distribution;

[0027] Perform feature restoration on the first compressed pollution data feature distribution with the second feature dimension quantity, and output the initial pollution data feature distribution with the second feature dimension quantity;

[0028] Perform an enhancement process on the feature dimension quantity of the initial pollution data feature distribution with the second feature dimension quantity, and output the intermediate pollution data feature distribution with the first feature dimension quantity. Perform feature restoration on the intermediate pollution data feature distribution with the first feature dimension quantity and the first compressed pollution data feature distribution with the first feature dimension quantity, and output the initial pollution data feature distribution with the first feature dimension quantity.

[0029] In some preferred embodiments, in the above industrial pollution source monitoring method, the step of respectively identifying multiple initial pollution data feature distributions with different feature dimension quantities from each of the pollution data sorting queues includes:

[0030] For each pollution data sorting queue among the multiple pollution data sorting queues, perform data feature mining on this pollution data sorting queue, and output the first pollution data feature distribution corresponding to this pollution data sorting queue;

[0031] Perform feature compression on the first pollution data feature distribution, and output the first compressed pollution data feature distribution with the first feature dimension quantity corresponding to the first pollution data feature distribution;

[0032] Perform a reduction process and feature compression on the feature dimension measure of the first compressed contaminated data feature distribution with the first feature dimension measure, output the first compressed contaminated data feature distribution with the second feature dimension measure, and stop continuing the reduction process and feature compression after the feature dimension measure of the output first compressed contaminated data feature distribution belongs to the first numerical value;

[0033] Perform a reduction process on the feature dimension measure of the first compressed contaminated data feature distribution with the first numerical value of feature dimension measures, and output the first compressed contaminated data feature distribution with the second numerical value of feature dimension measures;

[0034] Perform feature restoration on the first compressed contaminated data feature distribution with the second numerical value of feature dimension measures, and output the initial contaminated data feature distribution with the second numerical value of feature dimension measures, where the difference between the second numerical value and the first numerical value is equal to one;

[0035] Perform an enhancement process on the feature dimension measure of the initial contaminated data feature distribution with the second numerical value of feature dimension measures, output the intermediate contaminated data feature distribution with the first numerical value of feature dimension measures, then splice the feature distributions of the intermediate contaminated data feature distribution with the first numerical value of feature dimension measures and the first compressed contaminated data feature distribution with the first numerical value of feature dimension measures, output the spliced contaminated data feature distribution with the first numerical value of feature dimension measures, and then perform feature restoration on the spliced contaminated data feature distribution with the first numerical value of feature dimension measures to output the initial contaminated data feature distribution with the first numerical value of feature dimension measures, and, after finally outputting the initial contaminated data feature distribution with the first feature dimension measure, stop performing feature restoration, and the initial contaminated data feature distribution with the first feature dimension measure is the initial contaminated data feature distribution with the first feature dimension measure.

[0036] In some preferred embodiments, in the above industrial pollution source monitoring method, the step of performing a reduction process and feature compression on the first compressed contaminated data feature distribution with the first feature dimension measure to output the first compressed contaminated data feature distribution with the second feature dimension measure includes:

[0037] Perform a reduction process on the feature dimension measure of the first compressed contaminated data feature distribution with the first feature dimension measure to output the first compressed contaminated data feature distribution with the second feature dimension measure;

[0038] Perform feature partitioning on the first compressed contaminated data feature distribution to output multiple first sub-compressed contaminated data feature distributions, and each first sub-compressed contaminated data feature distribution includes multiple distribution parameters in the first compressed contaminated data feature distribution;

[0039] For each of the distribution parameters, according to the multiple distribution parameters included in the first sub-compressed pollution data feature distribution corresponding to the distribution parameter and the distribution information of the multiple distribution parameters, the distribution parameter is adjusted, and the distribution information is used to reflect the distribution relationship of the corresponding distribution parameter in the first sub-compressed pollution data feature distribution;

[0040] Based on the multiple distribution parameters adjusted according to one of the first sub-compressed pollution data feature distributions, a corresponding second sub-compressed pollution data feature distribution is formed;

[0041] According to the distribution relationship of the multiple first sub-compressed pollution data feature distributions in the first compressed pollution data feature distribution, the corresponding multiple second sub-compressed pollution data feature distributions are fused to output a corresponding first fusion feature distribution;

[0042] According to the first fusion feature distribution, a first compressed pollution data feature distribution with a second feature dimension is formed.

[0043] In some preferred embodiments, in the above industrial pollution source monitoring method, the step of, for each of the distribution parameters, adjusting the distribution parameter according to the multiple distribution parameters included in the first sub-compressed pollution data feature distribution corresponding to the distribution parameter and the distribution information of the multiple distribution parameters includes:

[0044] For each of the first sub-compressed pollution data feature distributions, each distribution parameter included in the first sub-compressed pollution data feature distribution is respectively fused with the distribution information corresponding to the distribution parameter, and multiple initial distribution features corresponding to the first sub-compressed pollution data feature distribution are output;

[0045] For each distribution parameter included in the first sub-compressed pollution data feature distribution, according to the feature matching degree between the initial distribution feature corresponding to the distribution parameter and the multiple initial distribution features, feature distribution splicing with splicing weights is performed on the multiple initial distribution features, and the output result is marked as the adjusted distribution parameter.

[0046] In some preferred embodiments, in the above industrial pollution source monitoring method, the step of comparing and analyzing the adjusted target pollution data set and the adjusted historical pollution data set to output second pollution change amplitude information, and then fusing the second pollution change amplitude information and the first pollution change amplitude information to form target pollution change amplitude information includes:

[0047] According to the corresponding industrial pollution source monitoring equipment and the internal moments within the corresponding time period, perform one-to-one correspondence processing of the pollution data between the adjusted target pollution data set and the adjusted historical pollution data set, so that each piece of adjusted industrial pollution data corresponds to a piece of adjusted historical industrial pollution data, so that each piece of adjusted industrial pollution data and the corresponding adjusted historical industrial pollution data have the same industrial pollution source monitoring equipment and internal moment within the time period, and the internal moment within the time period is used to reflect the sequence of collection times of the corresponding pollution data within the corresponding time period;

[0048] Calculate the data difference values between each group of corresponding adjusted industrial pollution data and adjusted historical industrial pollution data respectively, and then fuse the data difference values between each group of corresponding adjusted industrial pollution data and adjusted historical industrial pollution data to output the second pollution change amplitude information;

[0049] Fuse the second pollution change amplitude information and the first pollution change amplitude information to form the target pollution change amplitude information.

[0050] An embodiment of the present invention further provides an industrial pollution source monitoring system, which is applied to a pollution source monitoring server. The pollution source monitoring server is communicatively connected to a plurality of industrial pollution source monitoring equipment, and the plurality of industrial pollution source monitoring equipment are arranged in sequence to respectively collect pollution data for a corresponding industrial pollution area among a plurality of industrial pollution areas included in the industrial pollution source. The system includes:

[0051] A data comparison and analysis module, configured to compare and analyze the target pollution data set and the historical pollution data set to output the first pollution change amplitude information. The target pollution data set includes multiple pieces of industrial pollution data collected by each industrial pollution source monitoring equipment among the plurality of industrial pollution source monitoring equipment within the current time period, and the historical pollution data set includes multiple pieces of historical industrial pollution data collected by each industrial pollution source monitoring equipment among the plurality of industrial pollution source monitoring equipment within the previous time period;

[0052] A data adjustment module, configured to output an adjusted target pollution data set that matches the target pollution data set by performing data adjustment according to the target pollution data set, and then output an adjusted historical pollution data set that matches the historical pollution data set by performing data adjustment according to the historical pollution data set;

[0053] A change amplitude fusion module, configured to compare and analyze the adjusted target pollution data set and the adjusted historical pollution data set to output the second pollution change amplitude information, and then fuse the second pollution change amplitude information and the first pollution change amplitude information to form the target pollution change amplitude information.

[0054] An industrial pollution source monitoring method and system provided by an embodiment of the present invention compare and analyze a target pollution data set and a historical pollution data set to output first pollution change amplitude information. According to the target pollution data set, data adjustment is performed to output an adjusted target pollution data set that matches the target pollution data set. Then, according to the historical pollution data set, data adjustment is performed to output an adjusted historical pollution data set that matches the historical pollution data set. The adjusted target pollution data set and the adjusted historical pollution data set are compared and analyzed to output second pollution change amplitude information, and then the second pollution change amplitude information and the first pollution change amplitude information are fused to form target pollution change amplitude information. Through the above content, since not only the target pollution data set and the historical pollution data set are compared and analyzed to output the first pollution change amplitude information, but also the target pollution data set and the historical pollution data set are respectively adjusted, it is possible to compare and analyze the obtained adjusted target pollution data set and adjusted historical pollution data set to output the second pollution change amplitude information, and then fuse the two pollution change amplitude information to obtain the target pollution change amplitude information, making the basis for determining the target pollution change amplitude information more sufficient, so as to improve the reliability of pollution change amplitude monitoring to a certain extent, thereby improving the problem of low monitoring reliability in the prior art.

[0055] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following specifically enumerates preferred embodiments and, in conjunction with the accompanying drawings, makes a detailed description as follows. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 It is a structural block diagram of a pollution source monitoring server provided by an embodiment of the present invention; Figure 2 It is a schematic flowchart of each step included in the industrial pollution source monitoring method provided by an embodiment of the present invention; Figure 3 It is a schematic diagram of each module included in the industrial pollution source monitoring system provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0057] To make the objects, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Usually, the components of the embodiments of the present invention described and shown in the accompanying drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the present invention claimed, but only represents the selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0058] Referring to Figure 1 the content of, an embodiment of the present invention provides a pollution source monitoring server. Among them, the pollution source monitoring server may include a memory and a processor.

[0059] It should be understood that, in some exemplary embodiments, the memory and the processor are directly or indirectly electrically connected to achieve data transmission or interaction. For example, they may be electrically connected through one or more communication buses or signal lines. The memory may store at least one software functional module (computer program) that can exist in the form of software or firmware. The processor may be used to execute the executable computer program stored in the memory, thereby implementing the industrial pollution source monitoring method provided by the embodiment of the present invention.

[0060] It should be understood that, in some exemplary embodiments, the memory may be, but is not limited to, a random access memory (RAM), a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), etc. The processor may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), a system on chip (SoC), etc.; it may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0061] It should be understood that, in some exemplary embodiments, Figure 1 the structure shown is only schematic, and the pollution source monitoring server may further include more or fewer components than those shown in Figure 1 or have a different configuration from that shown in Figure 1 For example, it may include a communication unit for information interaction with other devices (such as industrial pollution source monitoring devices, where the industrial pollution source monitoring devices vary according to different industrial pollution sources, such as air pollution sources and water pollution sources, etc.).

[0062] It should be understood that in some exemplary embodiments, the pollution source monitoring server is communicatively connected to a plurality of industrial pollution source monitoring devices, which are arranged in sequence to respectively collect pollution data for a corresponding one of the plurality of industrial pollution areas included in the industrial pollution source (the plurality of industrial pollution areas are arranged in sequence).

[0063] Referring to Figure 2 the content of, an embodiment of the present invention further provides an industrial pollution source monitoring method, which can be applied to the above pollution source monitoring server. Among them, the method steps defined by the process related to the industrial pollution source monitoring method can be implemented by the pollution source monitoring server.

[0064] The following will Figure 2 elaborate in detail on the specific process shown.

[0065] Step S110, compare and analyze the target pollution data set and the historical pollution data set to output the first pollution change amplitude information.

[0066] In an embodiment of the present invention, the pollution source monitoring server can compare and analyze the target pollution data set and the historical pollution data set to output the first pollution change amplitude information. The target pollution data set includes multiple pieces of industrial pollution data collected by each of the plurality of industrial pollution source monitoring devices during the current time period, and the historical pollution data set includes multiple pieces of historical industrial pollution data (such as pollutant concentration data, etc.) collected by each of the plurality of industrial pollution source monitoring devices during the previous time period.

[0067] Step S120, based on the target pollution data set, output an adjusted target pollution data set that matches the target pollution data set through data adjustment, and then based on the historical pollution data set, output an adjusted historical pollution data set that matches the historical pollution data set through data adjustment.

[0068] In an embodiment of the present invention, the pollution source monitoring server can, based on the target pollution data set, output an adjusted target pollution data set that matches the target pollution data set through data adjustment, and then based on the historical pollution data set, output an adjusted historical pollution data set that matches the historical pollution data set through data adjustment (the way of performing data adjustment on the historical pollution data set can refer to the processing method of the target pollution data set described later).

[0069] Step S130: Compare and analyze the adjusted target pollution data set and the adjusted historical pollution data set to output second pollution change amplitude information, and then fuse the second pollution change amplitude information and the first pollution change amplitude information to form target pollution change amplitude information.

[0070] In an embodiment of the present invention, the pollution source monitoring server may compare and analyze the adjusted target pollution data set and the adjusted historical pollution data set to output second pollution change amplitude information, and then fuse the second pollution change amplitude information and the first pollution change amplitude information to form target pollution change amplitude information.

[0071] Through the foregoing content, since not only the target pollution data set and the historical pollution data set are compared and analyzed to output the first pollution change amplitude information, but also the target pollution data set and the historical pollution data set are respectively adjusted, it is possible to compare and analyze the obtained adjusted target pollution data set and the adjusted historical pollution data set to output second pollution change amplitude information, and then fuse the two pollution change amplitude information to obtain the target pollution change amplitude information, making the basis for determining the target pollution change amplitude information more sufficient, thereby improving the reliability of pollution change amplitude monitoring to a certain extent and thus improving the problem of low monitoring reliability in the prior art.

[0072] It should be understood that in some exemplary embodiments, the description of step S110 above can be understood through the following specific content:

[0073] According to the corresponding industrial pollution source monitoring equipment and the internal time moments within the corresponding time period, perform one-to-one correspondence processing of the pollution data of the target pollution data set and the historical pollution data set, so that each piece of industrial pollution data corresponds to a piece of historical industrial pollution data, so that there is the same industrial pollution source monitoring equipment and internal time moment within the time period between each piece of industrial pollution data and the corresponding historical industrial pollution data, and the internal time moment within the time period is used to reflect the sequence of the acquisition time of the corresponding pollution data within the corresponding time period;

[0074] Calculate the data difference value between each group of corresponding industrial pollution data and historical industrial pollution data (such as the absolute difference between industrial pollution data and historical industrial pollution data), and then fuse the data difference values between each group of corresponding industrial pollution data and historical industrial pollution data (such as performing mean calculation, etc.) to output the first pollution change amplitude information.

[0075] It should be understood that in some other exemplary embodiments, the description of step S110 above can be understood through the following specific content:

[0076] According to the target pollution dataset, a first pollution data sorting queue is formed, and then according to the historical pollution dataset, a second pollution data sorting queue is formed. The first pollution data sorting queue has a third number of rows and a fourth number of columns, and the second pollution data sorting queue has a third number of rows and a fourth number of columns. The third number is equal to the number of internal moments in the multiple time periods included in the time period, and the fourth number is equal to the number of the multiple industrial pollution source monitoring devices;

[0077] The first pollution data sorting queue is decomposed to form multiple first pollution data sorting sub-queues corresponding to the first pollution data sorting queue. Then, the second pollution data sorting queue is decomposed to form multiple second pollution data sorting sub-queues corresponding to the second pollution data sorting queue. The sum of the number of rows corresponding to the multiple first pollution data sorting sub-queues is equal to the third number (that is, the decomposition boundary is parallel to the row direction). The number of columns of each first pollution data sorting sub-queue is equal to the fourth number, and the number of columns of each second pollution data sorting sub-queue is equal to the fourth number;

[0078] For each first pollution data sorting sub-queue, a sliding window segmentation process is performed on the first pollution data sorting sub-queue (where the window size for the sliding window segmentation process is not limited, such as it can be 5*8, 2*3, 7*6, or 9*11, etc.) to output multiple first sliding window sub-queues corresponding to the first pollution data sorting sub-queue. And, the multiple first sliding window sub-queues corresponding to the first pollution data sorting sub-queue are compared one by one with the multiple first sliding window sub-queues corresponding to the next adjacent first pollution data sorting sub-queue, so as to perform an association process on the two identical first sliding window sub-queues between the first pollution data sorting sub-queue and the next adjacent first pollution data sorting sub-queue. And, according to each two first sliding window sub-queues with an association relationship, the multiple first sliding window sub-queues corresponding to the multiple first pollution data sorting sub-queues are connected respectively to form a first queue cascade path set. On each first queue cascade path included in the first queue cascade path set, there is an association relationship between any two adjacent first sliding window sub-queues (for example, if the first sliding window sequence a in the first pollution data sorting sub-queue A is the same as the first sliding window sequence b in the first pollution data sorting sub-queue B, then the first sliding window sequence a and the first sliding window sequence b are associated);

[0079] For each of the second contaminated data sorting sub - queues, perform a sliding window segmentation process on the second contaminated data sorting sub - queue to output a plurality of second sliding window sub - queues corresponding to the second contaminated data sorting sub - queue. And compare and analyze the plurality of second sliding window sub - queues with the plurality of second sliding window sub - queues corresponding to the next adjacent second contaminated data sorting sub - queue corresponding to the second contaminated data sorting sub - queue one by one, so as to perform an association process on two identical second sliding window sub - queues between the second contaminated data sorting sub - queue and the next two adjacent first contaminated data sorting sub - queues. And then, based on each pair of second sliding window sub - queues with an association relationship, connect the plurality of second sliding window sub - queues corresponding to the plurality of second contaminated data sorting sub - queues respectively to form a second queue cascade path set. On each second queue cascade path included in the second queue cascade path set, there is an association relationship between any two adjacent second sliding window sub - queues;

[0080] Extract a target first queue cascade path from the first queue cascade path set, and then extract a target second queue cascade path from the second queue cascade path set. The target first queue cascade path is the first queue cascade path with the largest screening coefficient in the first queue cascade path set, and the target second queue cascade path is the second queue cascade path with the largest screening coefficient in the second queue cascade path set. The screening coefficient is positively correlated with the number of sliding window sub - queues included in the corresponding queue cascade path, and the screening coefficient is negatively correlated with the degree of difference between the sub - queue distribution positions of the sliding window sub - queues included in the corresponding queue cascade path;

[0081] Determine a first target sub - queue distribution position according to the sub - queue distribution positions of each first sliding window sub - queue included in the target first queue cascade path (such as taking the union), and then determine a second target sub - queue distribution position according to the sub - queue distribution positions of each second sliding window sub - queue included in the target second queue cascade path (such as taking the union). Then, fuse (such as taking the union or taking the intersection, etc.) the first target sub - queue distribution position and the second target sub - queue distribution position to output the target sub - queue distribution position. And for each corresponding first contaminated data sorting sub - queue and second contaminated data sorting sub - queue (which can have the same row sorting in the first contaminated data sorting queue and the second contaminated data sorting queue), calculate the data difference value (such as the absolute difference between the industrial contaminated data and the historical industrial contaminated data at the corresponding positions) of the industrial contaminated data and the historical industrial contaminated data at the part corresponding to the target sub - queue distribution position between the first contaminated data sorting sub - queue and the second contaminated data sorting sub - queue;

[0082] Fuse the data difference values between each corresponding first contaminated data sorting sub-queue and second contaminated data sorting sub-queue (such as mean calculation), and output the first contaminated change amplitude information.

[0083] It should be understood that in some exemplary embodiments, the description of step S120 above can be understood through the following specific content:

[0084] Based on the target contaminated data set, form an ordered set of contaminated data. The ordered set of contaminated data includes multiple contaminated data sorting queues corresponding to multiple internal time periods of a time cycle. Each internal time period of the time cycle includes multiple internal time moments of the time cycle. At each internal time moment of the time cycle, each industrial pollution source monitoring device collects contaminated data for the corresponding industrial pollution area to form corresponding industrial contaminated data. Each contaminated data sorting queue includes a first number of rows and a second number of columns. The first number is equal to the number of internal time moments of the time cycle, and the second number is equal to the number of industrial pollution source monitoring devices. The multiple contaminated data sorting queues include a first contaminated data sorting queue, and the first contaminated data sorting queue is any one of the multiple contaminated data sorting queues.

[0085] Respectively identify initial contaminated data feature distributions with different numbers of feature dimensions (such as two-dimensional, three-dimensional, four-dimensional, etc.) from each contaminated data sorting queue (the manifestation forms of the initial contaminated data feature distribution and other feature distributions can be arbitrary. For example, they can be represented in the form of vectors), and then splice the initial contaminated data feature distributions corresponding to a contaminated data sorting queue to output the spliced contaminated data feature distribution corresponding to each contaminated data sorting queue (by splicing, the amount of information of the features can be increased).

[0086] According to the feature matching degree between the spliced contaminated data feature distribution of each contaminated data sorting queue and the spliced contaminated data feature distribution of the first contaminated data sorting queue (the feature matching degree can be used to reflect the association degree between the contaminated data sorting queue and the first contaminated data sorting queue), perform feature distribution splicing with splicing weights on the spliced contaminated data feature distributions of the multiple contaminated data sorting queues, and output the initial spliced contaminated data feature distribution corresponding to the first contaminated data sorting queue.

[0087] According to the initial spliced contaminated data feature distribution, perform reduction processing on the first contaminated data sorting queue to output the reduced first contaminated data sorting queue.

[0088] After successively processing each pollution data sorting queue in the multiple pollution data sorting queues as the first pollution data sorting queue to output the restored first pollution data sorting queue corresponding to each pollution data sorting queue, then according to the restored first pollution data sorting queue corresponding to each pollution data sorting queue, a adjusted target pollution data set matching the target pollution data set is combined and formed.

[0089] It should be understood that in some exemplary embodiments, the description of the step of splicing the feature distributions of multiple initial pollution data feature distributions corresponding to a pollution data sorting queue above and outputting the spliced pollution data feature distribution corresponding to each pollution data sorting queue can be understood through the following specific content:

[0090] For any pollution data sorting queue, according to the feature dimension quantity with the maximum value among the multiple feature dimension quantities, the feature dimension quantity of each other initial pollution data feature distribution other than the initial pollution data feature distribution corresponding to the feature dimension quantity with the maximum value in the multiple initial pollution data feature distributions corresponding to this pollution data sorting queue is adjusted to the feature dimension quantity with the maximum value (for example, it can be adjusted by interpolation), so as to form multiple adjusted initial pollution data feature distributions, and make the feature dimension quantity of each adjusted initial pollution data feature distribution equal to the feature dimension quantity with the maximum value;

[0091] Fuse the multiple adjusted initial pollution data feature distributions (the feature dimension quantities of the multiple adjusted initial pollution data feature distributions are the same, so they can be fused), and output the fused pollution data feature distribution corresponding to the multiple adjusted initial pollution data feature distributions;

[0092] Perform feature mining on the fused pollution data feature distribution (for example, it can be performed through a convolutional network. In this way, by first fusing the feature distributions and then further performing feature mining on the fused feature distributions, the accuracy of the output feature distribution can be higher), and output the spliced pollution data feature distribution corresponding to the pollution data sorting queue.

[0093] It should be understood that in some exemplary embodiments, the description of the step of respectively identifying multiple initial pollution data feature distributions with different feature dimension quantities from each pollution data sorting queue above can be understood through the following specific content:

[0094] For each of the multiple contaminated data sorting queues, perform data feature mining on the contaminated data sorting queue, and output the first contaminated data feature distribution corresponding to the contaminated data sorting queue (for example, the contaminated data sorting queue can be divided to form multiple contaminated data sorting sub-queues, and then, data feature mining is respectively performed on each contaminated data sorting sub-queue to output the sub-feature distribution of each contaminated data sorting sub-queue, and then, according to the queue distribution relationship of the multiple contaminated data sorting sub-queues in the contaminated data sorting queue, the sub-feature distributions corresponding to the multiple contaminated data sorting sub-queues are spliced to form the first contaminated data feature distribution of the contaminated data sorting queue);

[0095] Perform feature compression on the first contaminated data feature distribution (by performing feature compression, the correlation degree between the distribution parameters included in the first compressed contaminated data feature distribution can be improved to a certain extent, so that the first compressed contaminated data feature distribution can better represent the characteristics of the contaminated data included in the contaminated data sorting queue. In addition, as described above, the feature distribution can be a vector, for example, then the distribution parameter can be a vector coordinate value), and output the first compressed contaminated data feature distribution with the first feature dimension corresponding to the first contaminated data feature distribution;

[0096] Perform reduction processing on the feature dimension of the first compressed contaminated data feature distribution with the first feature dimension, and output the first compressed contaminated data feature distribution with the second feature dimension corresponding to the first contaminated data feature distribution;

[0097] Perform feature restoration on the first compressed contaminated data feature distribution with the second feature dimension (for example, the first compressed contaminated data feature distribution can be processed through a decoding network), and output the initial contaminated data feature distribution with the second feature dimension;

[0098] Perform enhancement processing on the feature dimension of the initial contaminated data feature distribution with the second feature dimension, output the intermediate contaminated data feature distribution with the first feature dimension, perform feature restoration on the intermediate contaminated data feature distribution with the first feature dimension and the first compressed contaminated data feature distribution with the first feature dimension, and output the initial contaminated data feature distribution with the first feature dimension.

[0099] It should be understood that in some other exemplary embodiments, the description of the step of respectively identifying multiple initial contaminated data feature distributions with different feature dimensions from each of the contaminated data sorting queues above can be understood through the following specific content:

[0100] For each of the multiple contaminated data sorting queues, perform data feature mining on the contaminated data sorting queue, and output the first contaminated data feature distribution corresponding to the contaminated data sorting queue;

[0101] Perform feature compression on the first contaminated data feature distribution, and output the first compressed contaminated data feature distribution with the first feature dimension quantity corresponding to the first contaminated data feature distribution;

[0102] Perform processing for reducing the feature dimension quantity and feature compression on the first compressed contaminated data feature distribution with the first feature dimension quantity, and output the first compressed contaminated data feature distribution with the second feature dimension quantity. After the feature dimension quantity of the output first compressed contaminated data feature distribution belongs to the first numerical quantity, stop continuing to perform processing for reducing the feature dimension quantity and feature compression (that is to say, in the above-described manner of forming the first compressed contaminated data feature distribution with the second feature dimension quantity, continue to obtain the first compressed contaminated data feature distribution with the next feature dimension quantity until the first compressed contaminated data feature distribution with the first numerical quantity of feature dimension quantities is obtained. For example, perform processing for reducing the feature dimension quantity and feature compression on the first compressed contaminated data feature distribution with the second feature dimension quantity to form the first compressed contaminated data feature distribution with the 3rd feature dimension quantity, then perform processing for reducing the feature dimension quantity and feature compression on the first compressed contaminated data feature distribution with the 3rd feature dimension quantity to form the first compressed contaminated data feature distribution with the 4th feature dimension quantity, then perform processing for reducing the feature dimension quantity and feature compression on the first compressed contaminated data feature distribution with the 4th feature dimension quantity to form the first compressed contaminated data feature distribution with the 5th feature dimension quantity);

[0103] Perform processing for reducing the feature dimension quantity on the first compressed contaminated data feature distribution with the first numerical quantity of feature dimension quantities, and output the first compressed contaminated data feature distribution with the second numerical quantity of feature dimension quantities;

[0104] Perform feature restoration on the first compressed contaminated data feature distribution with the second numerical quantity of feature dimension quantities, and output the initial contaminated data feature distribution with the second numerical quantity of feature dimension quantities, where the difference between the second numerical quantity and the first numerical quantity is equal to one;

[0105] Perform enhancement processing on the feature dimension measure of the initial contaminated data feature distribution with the second numerical value of the feature dimension measure, and output the intermediate contaminated data feature distribution with the first numerical value of the feature dimension measure. Then, splice the intermediate contaminated data feature distribution with the first numerical value of the feature dimension measure and the first compressed contaminated data feature distribution with the first numerical value of the feature dimension measure to output the spliced contaminated data feature distribution with the first numerical value of the feature dimension measure. Next, perform feature restoration on the spliced contaminated data feature distribution with the first numerical value of the feature dimension measure to output the initial contaminated data feature distribution with the first numerical value of the feature dimension measure. And, after finally outputting the initial contaminated data feature distribution with the first feature dimension measure, (referring to the above relevant description) stop performing feature restoration. The initial contaminated data feature distribution with the first feature dimension measure is the initial contaminated data feature distribution with the first feature dimension measure.

[0106] It should be understood that in some exemplary embodiments, the description of the step of performing reduction processing and feature compression on the feature dimension measure of the first compressed contaminated data feature distribution with the first feature dimension measure and outputting the first compressed contaminated data feature distribution with the second feature dimension measure in the above text can be understood through the following specific content:

[0107] Perform reduction processing on the feature dimension measure of the first compressed contaminated data feature distribution with the first feature dimension measure to output the first compressed contaminated data feature distribution with the second feature dimension measure;

[0108] Perform feature partitioning on the first compressed contaminated data feature distribution to output a plurality of first sub-compressed contaminated data feature distributions (the plurality of first sub-compressed contaminated data feature distributions can be spliced to form the first compressed contaminated data feature distribution), and each of the first sub-compressed contaminated data feature distributions includes a plurality of distribution parameters in the first compressed contaminated data feature distribution;

[0109] For each of the distribution parameters, adjust the distribution parameter according to the plurality of distribution parameters included in the first sub-compressed contaminated data feature distribution corresponding to the distribution parameter and the distribution information of the plurality of distribution parameters, and the distribution information is used to reflect the distribution relationship of the corresponding distribution parameter in the first sub-compressed contaminated data feature distribution;

[0110] Based on the adjusted plurality of distribution parameters of one of the first sub-compressed contaminated data feature distributions, form the corresponding second sub-compressed contaminated data feature distribution;

[0111] Fuse the corresponding multiple second sub-compressed pollution data feature distributions according to the distribution relationship of the multiple first sub-compressed pollution data feature distributions in the first compressed pollution data feature distribution, and output the corresponding first fused feature distribution;

[0112] According to the first fused feature distribution, form a first compressed pollution data feature distribution with a second feature dimension (for example, the first fused feature distribution can be directly marked as the first compressed pollution data feature distribution, or the first fused feature distribution can be linearly mapped, such as LinearProjection, to obtain the first compressed pollution data feature distribution).

[0113] It should be understood that in some exemplary embodiments, the above description of the step of adjusting each of the distribution parameters according to the multiple distribution parameters included in the first sub-compressed pollution data feature distribution corresponding to the distribution parameter and the distribution information of the multiple distribution parameters can be understood through the following specific content:

[0114] For each of the first sub-compressed pollution data feature distributions, fuse each distribution parameter included in the first sub-compressed pollution data feature distribution with the distribution information corresponding to the distribution parameter, and output multiple initial distribution features corresponding to the first sub-compressed pollution data feature distribution (for the convenience of fusing the distribution parameter and the corresponding distribution information, when the distribution parameter is represented by a vector, the distribution information can also be represented by a vector. In this way, the first sub-compressed pollution data feature distribution can be a vector formed by combining vectors corresponding to multiple distribution parameters. In this way, the distribution relationship of the distribution parameter can be fused into the distribution parameter, that is, the distribution relationship of the distribution parameter is also used as a feature to reflect the distribution parameter);

[0115] For each distribution parameter included in the first sub-compressed pollution data feature distribution, according to the feature matching degree between the initial distribution feature corresponding to the distribution parameter and the multiple initial distribution features, perform feature distribution splicing (i.e., weighted fusion processing) with splicing weights on the multiple initial distribution features, and then mark the output result as the adjusted distribution parameter.

[0116] It should be understood that in the above exemplary embodiments, the execution process of data adjustment on the target pollution data set through step S120 to output an adjusted target pollution data set matching the target pollution data set can be implemented by a pre-trained neural network model. The specific data processing flow of this neural network model can refer to the detailed content included in step S120 above. For example, the neural network model may include a convolutional network unit, a decoding network unit, etc. Among them, the training process of the neural network model can refer to existing related technologies. For example, an initial neural network model can be used to adjust an example pollution data set to output a corresponding estimated pollution data set, and then the data feature difference between the example pollution data set and the estimated pollution data set is calculated to output a corresponding data feature difference degree, and then this data feature difference degree is used as a network convergence evaluation value to update the initial neural network model to form a corresponding neural network model.

[0117] It should be understood that in some exemplary embodiments, the description of step S130 above can be understood through the following specific content:

[0118] According to the corresponding industrial pollution source monitoring equipment and the internal moments within the corresponding time period, one-to-one correspondence processing of pollution data is performed on the adjusted target pollution data set and the adjusted historical pollution data set, so that each adjusted industrial pollution data corresponds to an adjusted historical industrial pollution data, so that each adjusted industrial pollution data and the corresponding adjusted historical industrial pollution data have the same industrial pollution source monitoring equipment and internal moments within the time period. The internal moments within the time period are used to reflect the sequence of the collection time of the corresponding pollution data within the corresponding time period;

[0119] The data difference values between each group of corresponding adjusted industrial pollution data and adjusted historical industrial pollution data are calculated respectively (refer to the relevant description above), and then the data difference values between each group of corresponding adjusted industrial pollution data and adjusted historical industrial pollution data are fused (refer to the relevant description above) to output the second pollution change amplitude information.

[0120] The second pollution change amplitude information and the first pollution change amplitude information are fused (such as weighted summation, etc.) to form the target pollution change amplitude information.

[0121] Referring to Figure 3 the content of, an industrial pollution source monitoring system is further provided in an embodiment of the present invention, which can be applied to the above pollution source monitoring server. Among them, the industrial pollution source monitoring system may include a data comparison and analysis module, a data adjustment module, and a change amplitude fusion module.

[0122] It should be understood that in some exemplary embodiments, the data comparison and analysis module is configured to compare and analyze a target pollution data set and a historical pollution data set to output first pollution change amplitude information. The target pollution data set includes multiple pieces of industrial pollution data collected by each industrial pollution source monitoring device among the multiple industrial pollution source monitoring devices in the current time period, and the historical pollution data set includes multiple pieces of historical industrial pollution data collected by each industrial pollution source monitoring device among the multiple industrial pollution source monitoring devices in the previous time period;

[0123] It should be understood that in some exemplary embodiments, the data adjustment module is configured to, based on the target pollution data set, perform data adjustment to output an adjusted target pollution data set that matches the target pollution data set, and then, based on the historical pollution data set, perform data adjustment to output an adjusted historical pollution data set that matches the historical pollution data set.

[0124] It should be understood that in some exemplary embodiments, the change amplitude fusion module is configured to compare and analyze the adjusted target pollution data set and the adjusted historical pollution data set to output second pollution change amplitude information, and then fuse the second pollution change amplitude information and the first pollution change amplitude information to form target pollution change amplitude information.

[0125] In summary, an industrial pollution source monitoring method and system provided by the present invention compare and analyze a target pollution data set and a historical pollution data set to output first pollution change amplitude information. Based on the target pollution data set, data adjustment is performed to output an adjusted target pollution data set that matches the target pollution data set, and then, based on the historical pollution data set, data adjustment is performed to output an adjusted historical pollution data set that matches the historical pollution data set. The adjusted target pollution data set and the adjusted historical pollution data set are compared and analyzed to output second pollution change amplitude information, and then the second pollution change amplitude information and the first pollution change amplitude information are fused to form target pollution change amplitude information. Through the foregoing content, since not only the target pollution data set and the historical pollution data set are compared and analyzed to output the first pollution change amplitude information, but also the target pollution data set and the historical pollution data set are respectively adjusted, it is possible to compare and analyze the obtained adjusted target pollution data set and adjusted historical pollution data set to output the second pollution change amplitude information, and then fuse the two pollution change amplitude information to obtain the target pollution change amplitude information, making the basis for determining the target pollution change amplitude information more sufficient, so as to improve the reliability of pollution change amplitude monitoring to a certain extent, thereby improving the problem of low monitoring reliability in the prior art.

[0126] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. An industrial pollution source monitoring method, characterized in that, Applied to a pollution source monitoring server, the pollution source monitoring server is communicatively connected to a plurality of industrial pollution source monitoring devices, and the plurality of industrial pollution source monitoring devices are arranged and distributed in sequence to respectively collect pollution data for a corresponding one of a plurality of industrial pollution areas included in an industrial pollution source. The method includes: Comparatively analyzing a target pollution data set and a historical pollution data set to output first pollution change amplitude information. The target pollution data set includes a plurality of industrial pollution data collected by each of the plurality of industrial pollution source monitoring devices during a current time period, and the historical pollution data set includes a plurality of historical industrial pollution data collected by each of the plurality of industrial pollution source monitoring devices during a previous time period; Based on the target pollution data set, output an adjusted target pollution data set that matches the target pollution data set through data adjustment, and then based on the historical pollution data set, output an adjusted historical pollution data set that matches the historical pollution data set through data adjustment; Comparatively analyze the adjusted target pollution data set and the adjusted historical pollution data set to output second pollution change amplitude information, and then fuse the second pollution change amplitude information and the first pollution change amplitude information to form target pollution change amplitude information; The step of comparatively analyzing the adjusted target pollution data set and the adjusted historical pollution data set to output second pollution change amplitude information, and then fusing the second pollution change amplitude information and the first pollution change amplitude information to form target pollution change amplitude information includes: According to the corresponding industrial pollution source monitoring device and the internal moment of the corresponding time period, perform one-to-one correspondence processing on the pollution data of the adjusted target pollution data set and the adjusted historical pollution data set, so that each piece of adjusted industrial pollution data corresponds to a piece of adjusted historical industrial pollution data, so that each piece of adjusted industrial pollution data and the corresponding adjusted historical industrial pollution data have the same industrial pollution source monitoring device and internal moment of the time period. The internal moment of the time period is used to reflect the order of collection time of the corresponding pollution data within the corresponding time period; Calculate the data difference values between each group of corresponding adjusted industrial pollution data and adjusted historical industrial pollution data respectively, and then fuse the data difference values between each group of corresponding adjusted industrial pollution data and adjusted historical industrial pollution data to output second pollution change amplitude information; Fuse the second pollution change amplitude information and the first pollution change amplitude information to form target pollution change amplitude information.

2. The industrial pollution source monitoring method according to claim 1, characterized in that The step of comparatively analyzing the target pollution data set and the historical pollution data set to output first pollution change amplitude information includes: According to the corresponding industrial pollution source monitoring equipment and the internal moments within the corresponding time period, perform one-to-one correspondence processing of the pollution data between the target pollution data set and the historical pollution data set, so that each piece of industrial pollution data corresponds to a piece of historical industrial pollution data, so that each piece of the industrial pollution data and the corresponding historical industrial pollution data have the same industrial pollution source monitoring equipment and internal moments within the time period, and the internal moments within the time period are used to reflect the sequence of the collection time of the corresponding pollution data within the corresponding time period; Calculate the data difference values between each group of corresponding industrial pollution data and historical industrial pollution data respectively, and then fuse the data difference values between each group of corresponding industrial pollution data and historical industrial pollution data to output the first pollution change amplitude information.

3. The industrial pollution source monitoring method according to claim 1, characterized in that The step of outputting an adjusted target pollution data set that matches the target pollution data set by performing data adjustment according to the target pollution data set includes: According to the target pollution data set, form an ordered set of pollution data. The ordered set of pollution data includes multiple pollution data sorting queues corresponding to multiple internal time periods of the time period. Each internal time period of the time period includes multiple internal moments of the time period. At each internal moment of the time period, each industrial pollution source monitoring equipment collects pollution data for the corresponding industrial pollution area to form corresponding industrial pollution data. Each pollution data sorting queue includes a first number of rows and a second number of columns. The first number is equal to the number of internal moments of the time period, and the second number is equal to the number of industrial pollution source monitoring equipment. The multiple pollution data sorting queues include a first pollution data sorting queue, and the first pollution data sorting queue is any one of the multiple pollution data sorting queues; Respectively identify multiple initial pollution data feature distributions with different characteristic dimensions from each pollution data sorting queue, and then splice the multiple initial pollution data feature distributions corresponding to a pollution data sorting queue to output the spliced pollution data feature distribution corresponding to each pollution data sorting queue; According to the feature matching degree between the spliced pollution data feature distribution of each pollution data sorting queue and the spliced pollution data feature distribution of the first pollution data sorting queue, perform feature distribution splicing with splicing weights on the spliced pollution data feature distributions of the multiple pollution data sorting queues to output the initial spliced pollution data feature distribution corresponding to the first pollution data sorting queue; According to the initial spliced pollution data feature distribution, perform reduction processing on the first pollution data sorting queue to output the reduced first pollution data sorting queue; After successively processing each pollution data sorting queue in the multiple pollution data sorting queues as the first pollution data sorting queue to output the restored first pollution data sorting queue corresponding to each pollution data sorting queue, an adjusted target pollution data set matching the target pollution data set is then formed by combining according to the restored first pollution data sorting queue corresponding to each pollution data sorting queue.

4. The industrial pollution source monitoring method according to claim 3, characterized in that, The step of splicing the feature distributions of multiple initial pollution data feature distributions corresponding to a pollution data sorting queue and outputting the spliced pollution data feature distribution corresponding to each pollution data sorting queue includes: For any pollution data sorting queue, according to the feature dimension quantity with the maximum value among the multiple feature dimension quantities, adjust the feature dimension quantity of each other initial pollution data feature distribution other than the initial pollution data feature distribution corresponding to the feature dimension quantity with the maximum value in the multiple initial pollution data feature distributions corresponding to this pollution data sorting queue to the feature dimension quantity with the maximum value, so as to form multiple adjusted initial pollution data feature distributions, such that the feature dimension quantity of each adjusted initial pollution data feature distribution is equal to the feature dimension quantity with the maximum value; Fuse the multiple adjusted initial pollution data feature distributions and output the fused pollution data feature distribution corresponding to the multiple adjusted initial pollution data feature distributions; Perform feature mining on the fused pollution data feature distribution and output the spliced pollution data feature distribution corresponding to the pollution data sorting queue.

5. The industrial pollution source monitoring method according to claim 3, characterized in that, The step of respectively identifying multiple initial pollution data feature distributions with different feature dimension quantities from each pollution data sorting queue includes: For each pollution data sorting queue in the multiple pollution data sorting queues, perform data feature mining on this pollution data sorting queue and output the first pollution data feature distribution corresponding to this pollution data sorting queue; Perform feature compression on the first pollution data feature distribution and output the first compressed pollution data feature distribution with the first feature dimension quantity corresponding to the first pollution data feature distribution; Perform a reduction process on the feature dimension quantity of the first compressed pollution data feature distribution with the first feature dimension quantity and output the first compressed pollution data feature distribution with the second feature dimension quantity corresponding to the first pollution data feature distribution; Perform feature restoration on the first compressed pollution data feature distribution with the second feature dimension quantity and output the initial pollution data feature distribution with the second feature dimension quantity; Perform an enhancement process on the feature dimension quantity of the initial pollution data feature distribution with the second feature dimension quantity and output the intermediate pollution data feature distribution with the first feature dimension quantity. Perform feature restoration on the intermediate pollution data feature distribution with the first feature dimension quantity and the first compressed pollution data feature distribution with the first feature dimension quantity, and output the initial pollution data feature distribution with the first feature dimension quantity.

6. The industrial pollution source monitoring method according to claim 3, characterized in that, The step of identifying multiple initial pollution data feature distributions with different feature dimensions from each of the pollution data sorting queues respectively includes: For each of the multiple pollution data sorting queues, perform data feature mining on the pollution data sorting queue, and output the first pollution data feature distribution corresponding to the pollution data sorting queue; Perform feature compression on the first pollution data feature distribution, and output the first compressed pollution data feature distribution with the first feature dimension corresponding to the first pollution data feature distribution; Perform feature dimension reduction processing and feature compression on the first compressed pollution data feature distribution with the first feature dimension, output the first compressed pollution data feature distribution with the second feature dimension, and stop continuing with feature dimension reduction processing and feature compression after the feature dimension of the output first compressed pollution data feature distribution belongs to the first numerical value; Perform feature dimension reduction processing on the first compressed pollution data feature distribution with the first numerical value of feature dimensions, and output the first compressed pollution data feature distribution with the second numerical value of feature dimensions; Perform feature restoration on the first compressed pollution data feature distribution with the second numerical value of feature dimensions, and output the initial pollution data feature distribution with the second numerical value of feature dimensions, where the difference between the second numerical value and the first numerical value is equal to one; Perform feature dimension enhancement processing on the initial pollution data feature distribution with the second numerical value of feature dimensions, output the intermediate pollution data feature distribution with the first numerical value of feature dimensions, then splice the feature distributions of the intermediate pollution data feature distribution with the first numerical value of feature dimensions and the first compressed pollution data feature distribution with the first numerical value of feature dimensions, output the spliced pollution data feature distribution with the first numerical value of feature dimensions, then perform feature restoration on the spliced pollution data feature distribution with the first numerical value of feature dimensions, output the initial pollution data feature distribution with the first numerical value of feature dimensions, and, after finally outputting the initial pollution data feature distribution with the first feature dimension, stop performing feature restoration. The initial pollution data feature distribution with the first feature dimension is the initial pollution data feature distribution with the first feature dimension.

7. The industrial pollution source monitoring method according to claim 6, characterized in that, The step of performing feature dimension reduction processing and feature compression on the first compressed pollution data feature distribution with the first feature dimension, and outputting the first compressed pollution data feature distribution with the second feature dimension includes: Perform feature dimension reduction processing on the first compressed pollution data feature distribution with the first feature dimension, and output the first compressed pollution data feature distribution with the second feature dimension; Perform feature partitioning on the first compressed pollution data feature distribution, and output multiple first sub-compressed pollution data feature distributions. Each of the first sub-compressed pollution data feature distributions includes multiple distribution parameters in the first compressed pollution data feature distribution; For each of the said distribution parameters, according to the multiple distribution parameters included in the first sub-compressed pollution data characteristic distribution corresponding to the distribution parameter and the distribution information of the multiple distribution parameters, the distribution parameter is adjusted, and the distribution information is used to reflect the distribution relationship of the corresponding distribution parameter in the first sub-compressed pollution data characteristic distribution; Based on the multiple distribution parameters adjusted according to one of the first sub-compressed pollution data characteristic distributions, a corresponding second sub-compressed pollution data characteristic distribution is formed; According to the distribution relationship of the multiple first sub-compressed pollution data characteristic distributions in the first compressed pollution data characteristic distribution, the corresponding multiple second sub-compressed pollution data characteristic distributions are fused to output a corresponding first fusion characteristic distribution; According to the first fusion characteristic distribution, a first compressed pollution data characteristic distribution with a second characteristic dimension measure is formed.

8. The industrial pollution source monitoring method according to claim 7, characterized in that, The step of adjusting each of the said distribution parameters according to the multiple distribution parameters included in the first sub-compressed pollution data characteristic distribution corresponding to the distribution parameter and the distribution information of the multiple distribution parameters includes: For each of the first sub-compressed pollution data characteristic distributions, each distribution parameter included in the first sub-compressed pollution data characteristic distribution is respectively fused with the distribution information corresponding to the distribution parameter to output multiple initial distribution characteristics corresponding to the first sub-compressed pollution data characteristic distribution; For each distribution parameter included in the first sub-compressed pollution data characteristic distribution, according to the feature matching degree between the initial distribution characteristic corresponding to the distribution parameter and the multiple initial distribution characteristics, the multiple initial distribution characteristics are subjected to feature distribution splicing with splicing weights, and the output result is marked as the adjusted distribution parameter.

9. An industrial pollution source monitoring system, characterized in that, Applied to a pollution source monitoring server, the pollution source monitoring server is communicatively connected to multiple industrial pollution source monitoring devices, and the multiple industrial pollution source monitoring devices are arranged in sequence to respectively collect pollution data for a corresponding industrial pollution area among the multiple industrial pollution areas included in the industrial pollution source. The system includes: A data comparison and analysis module for comparing and analyzing a target pollution data set and a historical pollution data set to output first pollution change amplitude information. The target pollution data set includes multiple pieces of industrial pollution data collected by each of the multiple industrial pollution source monitoring devices during the current time period, and the historical pollution data set includes multiple pieces of historical industrial pollution data collected by each of the multiple industrial pollution source monitoring devices during the previous time period; A data adjustment module for adjusting data according to the target pollution data set to output an adjusted target pollution data set matching the target pollution data set, and then adjusting data according to the historical pollution data set to output an adjusted historical pollution data set matching the historical pollution data set; A variation range fusion module is used to compare and analyze the adjusted target pollution data set and the adjusted historical pollution data set to output second pollution variation range information, and then fuse the second pollution variation range information and the first pollution variation range information to form target pollution variation range information; The process of comparing and analyzing the adjusted target pollution data set and the adjusted historical pollution data set to output second pollution variation range information, and then fusing the second pollution variation range information and the first pollution variation range information to form target pollution variation range information includes: According to the corresponding industrial pollution source monitoring equipment and the internal moments within the corresponding time period, perform one-to-one correspondence processing of the pollution data in the adjusted target pollution data set and the adjusted historical pollution data set, so that each piece of adjusted industrial pollution data corresponds to one piece of adjusted historical industrial pollution data, so that each piece of adjusted industrial pollution data and the corresponding adjusted historical industrial pollution data have the same industrial pollution source monitoring equipment and internal moments within the time period, and the internal moments within the time period are used to reflect the sequence of the collection time of the corresponding pollution data within the corresponding time period; Calculate the data difference values between each group of corresponding adjusted industrial pollution data and adjusted historical industrial pollution data respectively, and then fuse the data difference values between each group of corresponding adjusted industrial pollution data and adjusted historical industrial pollution data to output second pollution variation range information; Fuse the second pollution variation range information and the first pollution variation range information to form target pollution variation range information.

Images