Water affair processing online management platform system and method based on artificial intelligence

Abstract

The invention discloses a water affair treatment online management platform system and method based on artificial intelligence, and belongs to the technical field of artificial intelligence and data processing, the operation data of sewage treatment equipment in the sewage affair treatment process are collected, and the operation data and timestamps are subjected to associated storage, so that effective traceability data can be formed; the water affair treatment process management is more convenient, and then the operation risk identification is performed by adopting the artificial intelligence model based on the operation data with the timestamp to obtain the risk identification result, so that the equipment operation risk in the water affair treatment process can be identified on line, and the management efficiency of the water affair treatment process is improved. And finally, on the basis of the risk identification result, emergency disposal operation and emergency early warning operation are executed, and the identified risk can be relieved or treated, so that the normal operation of the sewage water treatment process is ensured, and the sewage which does not reach the standard is prevented from flowing out.

Description

Technical Field

[0001] The present invention belongs to the field of artificial intelligence and data processing technology, and specifically relates to an artificial intelligence-based water treatment online management platform system and method. Background Art

[0002] Wastewater treatment is a vital component of environmental protection and public health, aiming to remove pollutants from wastewater and ensure it meets safe discharge or reuse standards. The treatment process typically includes physical, chemical, and biological methods, such as screening, grit settling, primary sedimentation, biological treatment, and disinfection. Physical methods primarily remove suspended matter and larger particles; chemical methods remove dissolved pollutants through the addition of chemicals; and biological methods utilize microorganisms to degrade organic matter. Modern wastewater treatment also incorporates technologies such as artificial intelligence and the Internet of Things to achieve intelligent management and optimized operations, improve treatment efficiency, reduce energy consumption, and ensure water quality. Wastewater treatment not only protects the water environment but also ensures the sustainable use of water resources. With accelerating urbanization and population growth, water treatment has become a vital component of urban infrastructure. Traditional water treatment systems often require regular manual inspections of equipment to ensure the proper operation of the wastewater treatment process. However, this can lead to inefficiencies and delayed risk detection, potentially resulting in substandard wastewater treatment quality. Summary of the Invention

[0003] The present invention provides an artificial intelligence-based online water treatment management platform system and method to solve the problems of low inspection efficiency and untimely risk discovery in the prior art.

[0004] On the one hand, the present invention provides an artificial intelligence-based water treatment online management platform system, comprising: an operation scheduling and supervision module, an early warning and prediction module, and an emergency response module; The operation scheduling and supervision module is used to collect the operation data of the sewage treatment equipment during the sewage treatment process and store the operation data in association with the timestamp; The early warning and prediction module is used to identify operational risks based on the operational data with timestamps using an artificial intelligence model to obtain risk identification results; The emergency response module is used to perform emergency response operations and emergency warning operations based on the risk identification results.

[0005] Furthermore, it also includes: a data storage module; The data storage module is used to store data generated during the operation of the operation scheduling and supervision module, the early warning and prediction module, and the emergency response module.

[0006] Furthermore, it also includes: a data online viewing module; The data online viewing module is used to allow staff to view the time-stamped operation data stored in the operation scheduling and supervision module, the risk identification results generated by the early warning and prediction module, and the operations performed by the emergency response module after the staff identity verification is passed.

[0007] Furthermore, the early warning and prediction module includes an artificial intelligence initialization submodule, an artificial intelligence deployment submodule, and a risk identification submodule; The artificial intelligence initialization submodule is used to build an artificial intelligence model; The artificial intelligence deployment submodule is used to optimize and deploy the artificial intelligence model generated by the artificial intelligence initialization submodule using a partition optimization algorithm to obtain a deployed artificial intelligence model; The risk identification submodule is used to construct the operating characteristics of the sewage treatment equipment on a periodic basis based on the operating data with timestamps, and schedule the deployed artificial intelligence model to identify the constructed operating characteristics to determine the risk identification results corresponding to each period; Among them, the risk identification results include the absence of risk or specific risk types.

[0008] Furthermore, a partition optimization algorithm is used to optimize the deployment of the artificial intelligence model generated by the artificial intelligence initialization submodule, including: Randomly initializing the parameters of the artificial intelligence model generated by the artificial intelligence initialization submodule to obtain a plurality of different parameter codes; wherein each parameter code includes all or part of the parameters to be optimized of the artificial intelligence model; For any parameter code, apply the parameters contained in the parameter code to the artificial intelligence model, use the historical operation characteristics as the input of the artificial intelligence model, and use the manual labels corresponding to the historical operation characteristics as the expected output to obtain the loss function value corresponding to the parameter code; wherein the historical operation characteristics and the manual labels corresponding to the historical operation characteristics are pre-stored data or data entered by staff; According to the loss function value corresponding to each parameter encoding, the parameter encoding with the smallest loss function value is determined as the optimal encoding; Based on the optimal code, the parameter code is partitioned, a search area corresponding to each parameter code is determined, and the parameter code after the partitioning process is obtained; Based on the location of the parameter code after the partitioning process, the location information of other parameter codes is used for reference, so that the parameter codes are searched in their own areas while searching in other areas to obtain the parameter codes after the area search; Based on the location of the parameter code after the area search, the average position information is used to search the parameter code towards the optimal area to obtain the parameter code after the optimal direction search; Performing an adaptive jump operation on the parameter code after the optimal direction search so that the parameter code jumps out of the area where it is located, and obtaining the parameter code after the jump search; Determine whether the maximum number of training times has been reached. If so, re-determine the optimal encoding based on the parameter encoding after the jump search, and use the parameters in the optimal encoding as the final parameters of the artificial intelligence model to deploy the artificial intelligence model and obtain the deployed artificial intelligence model. Otherwise, return to the step of obtaining the loss function value.

[0009] Furthermore, based on the optimal code, the parameter code is partitioned, a search area corresponding to each parameter code is determined, and the parameter code after the partitioning process is obtained, including: Averaging all parameter codes to determine the average parameter code; For any parameter code, determine the position information difference between the average parameter code and the parameter code to obtain the position information code; After the position information code is processed by using the position control variation parameter, the search area corresponding to the parameter code is determined in combination with the optimal parameter code to obtain the parameter code after partition processing.

[0010] Furthermore, based on the location of the parameter code after the partition processing, the location information of other parameter codes is used for reference, so that the parameter codes are searched in their respective areas while searching in other areas to obtain the parameter codes after the area search, including: Determining an adaptive inertia weight factor based on the current number of training times, and determining a nonlinear search angle control parameter according to the adaptive inertia weight factor; Determining a nonlinear search range control parameter based on the nonlinear search angle control parameter; Obtaining a first average position learning factor according to the nonlinear search angle control parameter and the nonlinear search range control parameter in combination with a sine function; According to the nonlinear search angle control parameter and the nonlinear search range control parameter, in combination with the cosine function, other parameter encoding position learning factors are obtained; The parameter codes after partitioning processing are arranged in descending order according to the loss function value, and the parameter codes after arrangement are searched according to the first average position learning factor and other parameter code position learning factors to obtain the parameter codes after the area search.

[0011] Furthermore, based on the location of the parameter code after the area search, the average position information is used to search the parameter code toward the optimal area, and the parameter code after the optimal direction search is obtained, including: Obtaining an optimal direction search step length control parameter according to the nonlinear search angle control parameter; Obtaining a second average position learning factor according to the nonlinear search angle control parameter and the optimal direction search step size control parameter in combination with a hyperbolic sine function; Obtaining an optimal position learning factor according to the nonlinear search angle control parameter and the optimal direction search step size control parameter in combination with a hyperbolic cosine function; According to the second average position learning factor, the optimal position learning factor and the optimal parameter code, the parameter code is searched towards the optimal area to obtain the parameter code after the optimal direction search.

[0012] Furthermore, an adaptive jump operation is performed on the parameter code after the optimal direction search so that the parameter code jumps out of the region where it is located, and the parameter code after the jump search is obtained, including: Determine the jumping ability control factor based on the current number of training sessions; According to the jump capability control factor, a jump operation is performed on the parameter code after the optimal direction search to obtain the parameter code after the jump search.

[0013] In another aspect, the present invention provides an artificial intelligence-based online water treatment management method, comprising: Collect the operating data of sewage treatment equipment during sewage treatment and store the operating data in association with timestamps; Based on the time-stamped operation data, an artificial intelligence model is used to identify operation risks and obtain risk identification results; Based on the risk identification results, emergency response operations and emergency warning operations are performed.

[0014] The present invention provides an artificial intelligence-based online management platform system and method for water treatment. By collecting the operating data of sewage treatment equipment during the sewage water treatment process and associating the operating data with timestamps for storage, effective traceability data can be formed, making the management of the water treatment process more convenient. Then, based on the operating data with timestamps, an artificial intelligence model is used to identify operating risks and obtain risk identification results. The equipment operating risks in the water treatment process can be identified online, thereby improving the management efficiency of the water treatment process. Finally, based on the risk identification results, emergency response operations and emergency warning operations are performed, and the identified risks can be alleviated or processed, thereby ensuring the normal progress of the sewage water treatment process and avoiding the outflow of sewage after substandard treatment. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0016] Figure 1 This is a structural diagram of an artificial intelligence-based water treatment online management platform system provided by the present invention.

[0017] Figure 2 This is a flow chart of an online water treatment management method based on artificial intelligence provided by the present invention.

[0018] The above drawings illustrate specific embodiments of the present invention, which will be described in more detail below. These drawings and the accompanying description are not intended to limit the scope of the present invention in any way, but rather to illustrate the concept of the present invention to those skilled in the art by reference to specific embodiments. DETAILED DESCRIPTION

[0019] Exemplary embodiments will be described in detail herein, examples of which are illustrated in the accompanying drawings. In the following description, when referring to the drawings, like numbers in different figures represent like or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all possible embodiments consistent with the present invention. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present invention, as detailed in the appended claims.

[0020] The embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0021] like Figure 1 As shown, the embodiment of the present invention provides an artificial intelligence-based water treatment online management platform system, including: an operation scheduling and supervision module 101, an early warning and prediction module 102, and an emergency response module 103; The operation scheduling and supervision module 101 is used to collect the operation data of the sewage treatment equipment during the sewage treatment process and store the operation data in association with the timestamp; Optionally, the operating data of the sewage treatment equipment includes at least: operating current, operating voltage and operating temperature. When the equipment is a motor-type equipment, it may also include operating speed and vibration frequency, so as to identify risks in the sewage treatment process.

[0022] For example, the primary wastewater pretreatment process primarily operates in the inlet pumphouse, which houses key equipment and facilities such as coarse screens, lift pumps, fine screens, sand-water separators, aerated grit chambers, primary sedimentation tanks, primary sedimentation sludge pumphouses, and supporting electrical, automation, and instrumentation equipment. During the primary pretreatment process, failure modes typically include damage to the coarse screen limit switch, resulting in screen failure and the inability to remove entangled debris; blockage in the inlet pump, preventing effective wastewater lifting; deformation or damage to the fine screen tines; malfunction of the Roots blower; failure of the sand-water separator's desander to extract slurry containing excessive organic matter and failing to operate; and damage to the crusher and substandard coolant in the sludge discharge system.

[0023] The sewage treatment equipment involved in the primary sewage pretreatment process basically belongs to the motor type, so the operating data may include operating speed, operating current, operating voltage and operating temperature.

[0024] The early warning and prediction module 102 is used to identify operational risks based on the operational data with timestamps using an artificial intelligence model to obtain risk identification results; Artificial intelligence models can be constructed using models such as convolutional neural networks, bidirectional long short-term memory networks, or BP neural networks. Then, based on timestamps, the operating data can be constructed as the input of the artificial intelligence model in cycles to achieve risk identification.

[0025] The risk identification result can be that there is no risk or there is a risk, or the risk identification result can be that there is no risk or a specific type of risk, both of which enable staff to quickly discover abnormalities and deal with them, ensuring the normal progress of the sewage and water treatment process.

[0026] The emergency response module 103 is used to perform emergency response operations and emergency warning operations based on the risk identification results.

[0027] In an embodiment of the present invention, the risk identification result is preferably that there is no risk or a specific risk type. Each specific risk type is pre-set with an emergency response operation, such as shutdown, reducing the operating speed, etc., and the emergency warning operation can be set to light warning operation and sound warning operation, so that staff can detect abnormal operating equipment in time.

[0028] In the embodiment of the present invention, it further includes: a data storage module 104; The data storage module 104 is used to store data generated during the operation of the operation scheduling and supervision module, the early warning and prediction module, and the emergency response module.

[0029] In the embodiment of the present invention, it further includes: a data online viewing module 105; The data online viewing module 105 is used to allow the staff to view the time-stamped operation data stored in the operation scheduling and supervision module, the risk identification results generated by the early warning and prediction module, and the operations performed by the emergency response module after the staff identity verification is passed.

[0030] In an embodiment of the present invention, the early warning and prediction module includes an artificial intelligence initialization submodule, an artificial intelligence deployment submodule, and a risk identification submodule; The artificial intelligence initialization submodule is used to build an artificial intelligence model; The artificial intelligence deployment submodule is used to optimize and deploy the artificial intelligence model generated by the artificial intelligence initialization submodule using a partition optimization algorithm to obtain a deployed artificial intelligence model; The risk identification submodule is used to construct the operating characteristics of the sewage treatment equipment on a periodic basis based on the operating data with timestamps, and schedule the deployed artificial intelligence model to identify the constructed operating characteristics to determine the risk identification results corresponding to each period; Among them, the risk identification results include the absence of risk or specific risk types.

[0031] In the process of optimizing and deploying artificial intelligence models using existing technologies, it is often easy to fall into local optimality, resulting in poor training results and ultimately inability to normally identify equipment operation risks. Therefore, an embodiment of the present invention proposes a partition optimization algorithm to improve the training effect and training accuracy of the algorithm, thereby improving the accuracy of equipment risk warning.

[0032] In an embodiment of the present invention, a partition optimization algorithm is used to optimize the deployment of the artificial intelligence model generated by the artificial intelligence initialization submodule, including: Randomly initializing the parameters of the artificial intelligence model generated by the artificial intelligence initialization submodule to obtain a plurality of different parameter codes; wherein each parameter code includes all or part of the parameters to be optimized of the artificial intelligence model; For example, when the artificial intelligence model is set as a convolutional neural network, some or all of its connection weights can be optimized. In addition to using random initialization, the chaotic mapping initialization method can also be used to generate multiple different parameter encodings.

[0033] For any parameter code, apply the parameters contained in the parameter code to the artificial intelligence model, use the historical operation features as the input of the artificial intelligence model, and use the manual labels corresponding to the historical operation features as the expected output to obtain the loss function value corresponding to the parameter code (such as the cross-entropy loss function); where the historical operation features and the manual labels corresponding to the historical operation features are pre-stored data or data entered by staff; According to the loss function value corresponding to each parameter encoding, the parameter encoding with the smallest loss function value is determined as the optimal encoding; Based on the optimal code, the parameter code is partitioned, a search area corresponding to each parameter code is determined, and the parameter code after the partitioning process is obtained; Based on the location of the parameter code after the partitioning process, the location information of other parameter codes is used for reference, so that the parameter codes are searched in their own areas while searching in other areas to obtain the parameter codes after the area search; Based on the location of the parameter code after the area search, the average position information is used to search the parameter code towards the optimal area to obtain the parameter code after the optimal direction search; Performing an adaptive jump operation on the parameter code after the optimal direction search so that the parameter code jumps out of the area where it is located, and obtaining the parameter code after the jump search; Determine whether the maximum number of training times has been reached. If so, re-determine the optimal encoding based on the parameter encoding after the jump search, and use the parameters in the optimal encoding as the final parameters of the artificial intelligence model to deploy the artificial intelligence model and obtain the deployed artificial intelligence model. Otherwise, return to the step of obtaining the loss function value.

[0034] In an embodiment of the present invention, based on the optimal code, the parameter code is partitioned, a search area corresponding to each parameter code is determined, and the parameter code after the partitioning process is obtained, including: All parameter codes are averaged to determine the average parameter code: ;in, Indicates the Average parameter encoding during the training process; the parameter of each dimension of the average parameter encoding is the average value of all parameters of the same dimension; For any parameter encoding, determine the position information difference between the average parameter encoding and the parameter encoding, and obtain the position information encoding as: ;in, Indicates the During the training n parameter encoding, n =1,2,…,N, N represents the total number of parameter codes, Indicates parameter encoding Corresponding position information code; After processing the position information encoding using the position control change parameter, the search area corresponding to the parameter encoding is determined in combination with the optimal parameter encoding, and the parameter encoding after partition processing is obtained as follows: ;in, represents the optimal parameter encoding, Indicates the position control change parameter, represents the first random number between (0,1), Indicates the parameter encoding after partition processing .

[0035] The partition processing operation provided by the embodiment of the present invention can enable each parameter code to select a search area around the optimal parameter code (i.e., centered on the optimal parameter code), which can greatly improve the search speed of the algorithm. At the same time, it draws on the position information of the average parameter code and can maintain diversity in the search process.

[0036] In the embodiment of the present invention, based on the location of the parameter code after the partition processing, the location information of other parameter codes is referenced, so that the parameter codes are searched in their respective areas while searching in other areas to obtain the parameter codes after the area search, including: Based on the current number of training times, the adaptive inertia weight factor is determined as: ;in, represents the adaptive inertia weight factor, Indicates the current number of training times. Indicates the preset maximum number of training times, Indicates the preset maximum value of the adaptive inertia weight factor, Indicates the preset minimum value of the adaptive inertia weight factor; The nonlinear search angle control parameter is determined according to the adaptive inertia weight factor: ;in, Indicates the The nonlinear search angle control parameters corresponding to the parameter encoding after the mth partition processing in the training process are: represents pi, Represents the second random number between (0,1); Based on the nonlinear search angle control parameter, the nonlinear search range control parameter is determined as: ;in, Indicates the The nonlinear search angle control parameters corresponding to the parameter encoding after the mth partition processing in the training process are: represents the third random number between (0,1), represents the spiral trajectory control factor; According to the nonlinear search angle control parameter and the nonlinear search range control parameter, combined with the sine function, the first average position learning factor is obtained as follows: ;in, represents the first mean position learning factor, Indicates the maximum value of the product of the nonlinear search angle control parameter and the nonlinear search range control parameter in the parameter encoding after all partition processing, Indicates the During the training k The nonlinear search angle control parameters corresponding to the parameter encoding after partition processing are Indicates the During the training k The parameter encoding after each partition processing corresponds to the nonlinear search angle control parameter, and sin represents the sine function; According to the nonlinear search angle control parameter and the nonlinear search range control parameter, combined with the cosine function, other parameter encoding position learning factors are obtained as follows: ;in, represents other parameter encoding position learning factors, cos represents the cosine function; Arrange the parameter codes after partitioning in descending order according to the loss function value, and search the parameter codes after arrangement according to the first average position learning factor and other parameter code position learning factors to obtain the parameter codes after regional search: ;in, Indicates the During the training m The parameter encoding after permutation processing, m =1,2,…,N, The parameter encoding after the area search is , Indicates the During the training m +1 parameter encoding after permutation processing, and when m is equal to N, Set to any one of the parameter codes after the N / 2th to Nth permutation processes.

[0037] The regional search method provided by the embodiment of the present invention allows each parameter code to search in its own search area based on the relationship between itself and the average position of the population. At the same time, it can learn the position information of other parameter codes to form a chain spiral search route. All parameter codes collaborate in the search and can form a mesh search path in the search space, further improving the ability to search for the global optimal solution.

[0038] In the embodiment of the present invention, based on the location of the parameter code after the area search, the average position information is used to search the parameter code toward the optimal area, and the parameter code after the optimal direction search is obtained, including: According to the nonlinear search angle control parameter, the optimal direction search step length control parameter is obtained as: ;in, Indicates the During the training j The parameter encoding after the area search corresponds to the nonlinear search angle control parameter, Indicates the During the training j The optimal direction search step length control parameter corresponding to the parameter encoding after the area search, j =1,2,…,N; According to the nonlinear search angle control parameter and the optimal direction search step size control parameter, combined with the hyperbolic sine function, the second average position learning factor is obtained as follows: ;in, represents the second mean position learning factor, Indicates the maximum value of the product of the nonlinear search angle control parameter and the optimal direction search step length control parameter in the parameter encoding after the area search. Indicates the During the training k The parameter encoding after the area search corresponds to the optimal direction search step control parameter, and sinh represents the hyperbolic sine function; According to the nonlinear search angle control parameter and the optimal direction search step size control parameter, combined with the hyperbolic cosine function, the optimal position learning factor is obtained as follows: ;in, represents the optimal position learning factor, cosh represents the hyperbolic cosine function; According to the second average position learning factor, the optimal position learning factor and the optimal parameter code, the parameter code is searched towards the optimal area, and the parameter code after the optimal direction search is obtained as follows: ;in, Indicates the During the training j Parameter encoding after area search, represents the optimal parameter encoding, Indicates the parameter encoding after the optimal direction search , represents the fourth random number between (0,1), represents the fifth random number between (0,1), Represents the sixth random number between (0,1).

[0039] The search in the optimal direction provided by the embodiment of the present invention can enable all parameter codes to search for the optimal position, which helps to improve the search speed of the algorithm. At the same time, it also integrates the average position information, which can enable parameter codes with poor positions to move quickly, thereby ensuring the training efficiency of the algorithm.

[0040] In an embodiment of the present invention, an adaptive jump operation is performed on the parameter code after the optimal direction search so that the parameter code jumps out of the region where the parameter code is located, and the parameter code after the jump search is obtained, including: Based on the current number of training sessions, the jumping ability control factor is determined as: ;in, represents the jumping ability control factor, e represents the natural constant, Represents the seventh random number between (0,1), represents pi; According to the jump capability control factor, a jump operation is performed on the parameter code after the optimal direction search, and the parameter code after the jump search is obtained as follows: ;in, Indicates the During the training h Parameter encoding after the optimal direction search, Indicates the parameter code after the jump search , h =1,2,…,N, represents the eighth random number between (0,1), represents the ninth random number between (0,1), represents the tenth random number between (0,1), represents the eleventh random number between (0,1), represents a parameter code randomly selected from the parameter codes after the optimal direction search, Represents a randomly generated new parameter code.

[0041] Optionally, an annealing simulation algorithm or a greedy principle may be used to control the jump operation process, thereby further ensuring the convergence speed of the algorithm.

[0042] The jump search provided by the embodiment of the present invention can effectively assist the algorithm to jump out of the local optimal solution, and adaptively reduce the jumping ability in the later stage of the algorithm to ensure further convergence of the algorithm, thereby improving the training effect of the algorithm. Combined with the previous search strategies, it can greatly improve the search ability and search accuracy of the global optimal solution, and ultimately improve the fault warning capability.

[0043] The present invention provides an artificial intelligence-based online management platform system for water treatment. By collecting the operating data of sewage treatment equipment during the sewage water treatment process and storing the operating data in association with a timestamp, effective traceability data can be formed, making the management of the water treatment process more convenient. Then, based on the operating data with a timestamp, an artificial intelligence model is used to identify operating risks and obtain risk identification results. The operating risks of equipment in the water treatment process can be identified online, thereby improving the management efficiency of the water treatment process. Finally, based on the risk identification results, emergency response operations and emergency warning operations are performed, which can alleviate or process the identified risks, thereby ensuring the normal progress of the sewage water treatment process and avoiding the outflow of sewage after substandard treatment.

[0044] like Figure 2 As shown, the present invention provides an online management method for water treatment based on artificial intelligence, comprising: S201. Collecting operation data of sewage treatment equipment during sewage treatment, and storing the operation data in association with a timestamp; S202. Based on the time-stamped operation data, an artificial intelligence model is used to identify operation risks and obtain risk identification results. S203: Based on the risk identification result, perform emergency response operations and emergency warning operations.

[0045] An artificial intelligence-based online management method for water treatment provided in an embodiment of the present invention can be applied to the above-mentioned system. Its principles and beneficial effects are similar and will not be described in detail here.

[0046] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0047] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0048] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0049] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0050] Those skilled in the art will understand that all or part of the steps in implementing the above facts and methods can be completed by instructing relevant hardware through a program, and the program involved or the program can be stored in a computer-readable storage medium. When the program is executed, it includes the following steps: the corresponding method steps are then brought out, and the storage medium can be ROM / RAM, a disk, an optical disk, etc.

[0051] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. An artificial intelligence-based water treatment online management platform system, characterized by: include: Operation scheduling and supervision module, early warning and prediction module, and emergency response module; The operation scheduling and supervision module is used to collect the operation data of the sewage treatment equipment during the sewage treatment process and store the operation data in association with the timestamp; The early warning and prediction module is used to identify operational risks based on the operational data with timestamps using an artificial intelligence model to obtain risk identification results; The emergency response module is used to perform emergency response operations and emergency warning operations based on the risk identification results.

2. The artificial intelligence-based water treatment online management platform system according to claim 1 is characterized in that: Also includes: Data storage module; The data storage module is used to store data generated during the operation of the operation scheduling and supervision module, the early warning and prediction module, and the emergency response module.

3. The artificial intelligence-based water treatment online management platform system according to claim 1 is characterized in that: Also includes: Data online viewing module; The data online viewing module is used to allow staff to view the time-stamped operation data stored in the operation scheduling and supervision module, the risk identification results generated by the early warning and prediction module, and the operations performed by the emergency response module after the staff identity verification is passed.

4. The artificial intelligence-based water treatment online management platform system according to claim 1 is characterized in that: The early warning and prediction module includes an artificial intelligence initialization submodule, an artificial intelligence deployment submodule, and a risk identification submodule; The artificial intelligence initialization submodule is used to build an artificial intelligence model; The artificial intelligence deployment submodule is used to optimize and deploy the artificial intelligence model generated by the artificial intelligence initialization submodule using a partition optimization algorithm to obtain a deployed artificial intelligence model; The risk identification submodule is used to construct the operating characteristics of the sewage treatment equipment on a periodic basis based on the operating data with timestamps, and schedule the deployed artificial intelligence model to identify the constructed operating characteristics to determine the risk identification results corresponding to each period; Among them, the risk identification results include the absence of risk or specific risk types.

5. The artificial intelligence-based water treatment online management platform system according to claim 4 is characterized in that: A partition optimization algorithm is used to optimize and deploy the artificial intelligence model generated by the artificial intelligence initialization submodule, including: Randomly initializing the parameters of the artificial intelligence model generated by the artificial intelligence initialization submodule to obtain a plurality of different parameter codes; wherein each parameter code includes all or part of the parameters to be optimized of the artificial intelligence model; For any parameter code, apply the parameters contained in the parameter code to the artificial intelligence model, use the historical operation characteristics as the input of the artificial intelligence model, and use the manual labels corresponding to the historical operation characteristics as the expected output to obtain the loss function value corresponding to the parameter code; wherein the historical operation characteristics and the manual labels corresponding to the historical operation characteristics are pre-stored data or data entered by staff; According to the loss function value corresponding to each parameter encoding, the parameter encoding with the smallest loss function value is determined as the optimal encoding; Based on the optimal code, the parameter code is partitioned, a search area corresponding to each parameter code is determined, and the parameter code after the partitioning process is obtained; Based on the location of the parameter code after the partitioning process, the location information of other parameter codes is used for reference, so that the parameter codes are searched in their own areas while searching in other areas to obtain the parameter codes after the area search; Based on the location of the parameter code after the area search, the average position information is used to search the parameter code towards the optimal area to obtain the parameter code after the optimal direction search; Performing an adaptive jump operation on the parameter code after the optimal direction search so that the parameter code jumps out of the area where it is located, and obtaining the parameter code after the jump search; Determine whether the maximum number of training times has been reached. If so, re-determine the optimal encoding based on the parameter encoding after the jump search, and use the parameters in the optimal encoding as the final parameters of the artificial intelligence model to deploy the artificial intelligence model and obtain the deployed artificial intelligence model. Otherwise, return to the step of obtaining the loss function value.

6. The artificial intelligence-based water treatment online management platform system according to claim 5 is characterized in that: Based on the optimal code, the parameter code is partitioned, a search area corresponding to each parameter code is determined, and the parameter code after the partitioning process is obtained, including: Averaging all parameter codes to determine the average parameter code; For any parameter code, determine the position information difference between the average parameter code and the parameter code to obtain the position information code; After the position information code is processed by using the position control variation parameter, the search area corresponding to the parameter code is determined in combination with the optimal parameter code to obtain the parameter code after partition processing.

7. The artificial intelligence-based water treatment online management platform system according to claim 6 is characterized in that: Based on the location of the parameter code after partition processing, and drawing on the location information of other parameter codes, the parameter codes are searched in their respective areas while searching in other areas to obtain the parameter codes after the area search, including: Determining an adaptive inertia weight factor based on the current number of training times, and determining a nonlinear search angle control parameter according to the adaptive inertia weight factor; Determining a nonlinear search range control parameter based on the nonlinear search angle control parameter; Obtaining a first average position learning factor according to the nonlinear search angle control parameter and the nonlinear search range control parameter in combination with a sine function; According to the nonlinear search angle control parameter and the nonlinear search range control parameter, in combination with the cosine function, other parameter encoding position learning factors are obtained; The parameter codes after partitioning processing are arranged in descending order according to the loss function value, and the parameter codes after arrangement are searched according to the first average position learning factor and other parameter code position learning factors to obtain the parameter codes after the area search.

8. The artificial intelligence-based water treatment online management platform system according to claim 7 is characterized in that: Based on the location of the parameter code after the area search, the average position information is used to search the parameter code towards the optimal area, and the parameter code after the optimal direction search is obtained, including: Obtaining an optimal direction search step length control parameter according to the nonlinear search angle control parameter; Obtaining a second average position learning factor according to the nonlinear search angle control parameter and the optimal direction search step size control parameter in combination with a hyperbolic sine function; Obtaining an optimal position learning factor according to the nonlinear search angle control parameter and the optimal direction search step size control parameter in combination with a hyperbolic cosine function; According to the second average position learning factor, the optimal position learning factor and the optimal parameter code, the parameter code is searched towards the optimal area to obtain the parameter code after the optimal direction search.

9. The artificial intelligence-based water treatment online management platform system according to claim 8 is characterized in that: An adaptive jump operation is performed on the parameter code after the optimal direction search to make the parameter code jump out of the region where it is located, and the parameter code after the jump search is obtained, including: Determine the jumping ability control factor based on the current number of training sessions; According to the jump capability control factor, a jump operation is performed on the parameter code after the optimal direction search to obtain the parameter code after the jump search.

10. An online management method for water treatment based on artificial intelligence, characterized in that: include: Collect the operating data of sewage treatment equipment during sewage treatment and store the operating data in association with timestamps; Based on the time-stamped operation data, an artificial intelligence model is used to identify operation risks and obtain risk identification results; Based on the risk identification results, emergency response operations and emergency warning operations are performed.

Images