Intelligent water affair system and intelligent control method applied by same

Abstract

The invention provides an intelligent water affair system and an intelligent control method applied by the intelligent water affair system, belongs to the field of intelligent water affair, and is used for solving the problem that the aeration air quantity of an aerator in an aerobic tank is difficult to accurately control in the sewage treatment process in the related technology. The adjustment amount of the aeration air volume and the application duration of the adjusted aeration air volume are intelligently analyzed based on the water volume, the water temperature, the dissolved oxygen amount, the internal reflux amount, the external reflux amount, the water quality index and the aeration air volume of the aerobic tank in a preset number of unit durations before the current moment, so that the adjusted aeration air volume meets the actual demand of sewage treatment. Whether the aeration air volume needs to be adjusted or not does not need to be judged in real time, accurate control over the aeration air volume in the aerobic tank in the sewage treatment process is achieved, and it is guaranteed that the treated water quality reaches the standard.

Description

Technical Field

[0001] The present application relates to the field of smart water services, and in particular to a smart water service system and an intelligent control method for its application. Background Art

[0002] Wastewater treatment includes a series of production processes such as aeration, denitrification, ozone, and coagulation. In order to ensure that the treated water quality meets the standards, it is necessary to continuously change the input of resources such as oxygen, carbon source, ozone, and flocculants according to the actual situation. This process is generally implemented by a smart water system. Smart water generally uses a series of monitoring equipment to determine the situation of sewage treatment, and based on feedback regulation and other regulatory mechanisms, determines the dosage of various resources used for sewage treatment. However, the actual situation of sewage treatment is relatively complicated, and it is difficult to accurately meet the actual resource needs of sewage treatment through feedback regulation and other regulatory mechanisms. Summary of the invention

[0003] The present application provides a smart water system and an intelligent control method for its application, which can be applied to the smart water system, and is conducive to accurately controlling the aeration air volume of the aerator in the aerobic tank during sewage treatment, so that the aeration air volume meets the actual needs of sewage treatment, so as to ensure that the treated water quality meets the standards.

[0004] In a first aspect, the present application provides an intelligent control method for a smart water system. The method comprises: Obtain the water volume, water temperature, dissolved oxygen content, internal return flow, external return flow, water quality indicators and aeration air volume of the aerobic pool within a preset number of unit time before the current time, wherein the water quality indicators include chemical oxygen demand, biochemical oxygen demand, ammonia nitrogen content, phosphorus content, and pH; The water volume, water temperature, dissolved oxygen content, internal recirculation volume, external recirculation volume, water quality index and aeration air volume of the aerobic pool within a preset number of unit time periods before the current moment are input into a pre-constructed aeration scheme analysis model to obtain the adjustment amount of the aeration air volume and the application time of the adjusted aeration air volume. The aeration scheme analysis model includes an input layer, multiple hidden layers and an output layer. The input layer includes input nodes that correspond one to one with the water volume, water temperature, dissolved oxygen content, internal recirculation volume, external recirculation volume, water quality index and aeration air volume of the aerobic pool within a preset number of unit time periods before the current moment. The output layer includes two output nodes for outputting the adjustment amount of the aeration air volume and the application time of the adjusted aeration air volume.

[0005] By adopting the above technical scheme, a pre-constructed aeration scheme analysis model can be used to intelligently analyze the adjustment amount of the aeration air volume and the application time of the adjusted aeration air volume based on the water volume, water temperature, dissolved oxygen content, internal recirculation volume, external recirculation volume, water quality indicators and aeration air volume in the aerobic pool within a preset number of unit time periods before the current moment, so that the adjusted aeration air volume can meet the actual needs of sewage treatment, and there is no need to judge in real time whether the aeration air volume needs to be adjusted. Accurate control of the aeration air volume in the aerobic pool during sewage treatment is achieved, which is conducive to ensuring that the treated water quality meets the standards.

[0006] Furthermore, the aeration scheme analysis model also includes a fluctuation value calculation sub-model and a change rate calculation sub-model; The fluctuation value calculation submodel and the change rate calculation submodel are both connected to the input nodes of the dissolved oxygen of a preset number of unit time lengths in the input layer, so that the fluctuation value calculation submodel outputs the dissolved oxygen fluctuation value, and the change rate calculation submodel outputs the dissolved oxygen change rate, the dissolved oxygen fluctuation value is positively correlated to the standard deviation of the dissolved oxygen in the aerobic pool within the preset number of unit time lengths and / or the absolute value of the difference between the dissolved oxygen in any two adjacent unit time lengths of the aerobic pool, and the dissolved oxygen change rate is positively correlated to the result of subtracting the dissolved oxygen in the aerobic pool in the latter of the two adjacent unit time lengths from the dissolved oxygen in the former; The hidden layer includes a first special neuron, and the first special neuron is used to input the output of each neuron in the previous hidden layer and the dissolved oxygen fluctuation value and the dissolved oxygen change rate.

[0007] Furthermore, the fluctuation value calculation sub-model includes: Assume that the preset number is n, and the dissolved oxygen content of the i-th unit time is , then the average value of dissolved oxygen , standard deviation of dissolved oxygen ; The sum of the absolute values ​​of the differences in dissolved oxygen at adjacent time steps ; The dissolved oxygen fluctuation value , where and are the pre-acquired standard deviation fluctuation coefficient and step difference fluctuation coefficient respectively.

[0008] Furthermore, the change rate calculation sub-model includes: Assume that the preset number is n, and the dissolved oxygen content of the i-th unit time is , calculate the average value of the difference between adjacent time steps of dissolved oxygen , the rate of change of dissolved oxygen is equal to the average value of the difference between adjacent time steps of dissolved oxygen .

[0009] Furthermore, the aeration scheme analysis model also includes a water quality fluctuation analysis model and a fluctuation trend analysis model; The water quality fluctuation analysis model and the fluctuation trend analysis model are both connected to the input nodes of the water quality indicators of a preset number of unit time lengths in the input layer, so that the water quality fluctuation analysis sub-model outputs water quality fluctuation data, and the fluctuation trend analysis sub-model outputs fluctuation trend data, the water quality fluctuation data is positively correlated to the standard deviation of the chemical oxygen demand, and / or biochemical oxygen demand, and / or ammonia nitrogen content, and / or phosphorus content, and / or pH within the preset number of unit time lengths, and the fluctuation trend data is positively correlated to the change rate of the chemical oxygen demand, and / or biochemical oxygen demand, and / or ammonia nitrogen content, and / or phosphorus content, and / or pH within the preset number of unit time lengths; The hidden layer includes a second special neuron, which is used to input the output of each neuron in the previous hidden layer and the water quality fluctuation data and fluctuation trend data.

[0010] Furthermore, the water quality fluctuation analysis sub-model includes: Assuming the preset number is n, the chemical oxygen demand, biochemical oxygen demand, ammonia nitrogen content, phosphorus content, and pH value of the i-th unit time are respectively , , , , , calculate the average values ​​of chemical oxygen demand, biochemical oxygen demand, ammonia nitrogen content, phosphorus content, and pH, respectively. , , , , , Then calculate the standard deviation of chemical oxygen demand, biochemical oxygen demand, ammonia nitrogen content, phosphorus content, and pH. , , , , , After calculating the fluctuation trend data .

[0011] Furthermore, the fluctuation trend analysis sub-model includes: Assuming the preset number is n, the chemical oxygen demand, biochemical oxygen demand, ammonia nitrogen content, phosphorus content, and pH value of the i-th unit time are respectively , , , , ; Linear regression model was used for chemical oxygen demand, biochemical oxygen demand, ammonia nitrogen content, phosphorus content, and pH and the least squares method to fit a straight line, where y is the chemical oxygen demand, or the biochemical oxygen demand, or the ammonia nitrogen content, or the phosphorus content, or the pH value, and x is the time in which the unit time is located; The fitting lines of chemical oxygen demand, biochemical oxygen demand, ammonia nitrogen content, phosphorus content, and pH are, , , , , , Determine the slopes of the fitted lines for chemical oxygen demand, biochemical oxygen demand, ammonia nitrogen content, phosphorus content, and pH, and calculate the fluctuation trend data based on the slopes. , In the formula, They are chemical oxygen demand, biochemical oxygen demand, ammonia nitrogen content, phosphorus content, and pH fluctuation trend attention coefficient.

[0012] In a second aspect, the present application provides a smart water system, wherein the system comprises a controller for controlling the adjustment amount of aeration air volume in an aerobic pool and the application duration of the adjusted aeration air volume, and the controller is configured to execute any one of the methods described in the first aspect above.

[0013] In summary, this application at least has the following beneficial effects: Provided is a smart water system and an intelligent control method for its application, which can intelligently analyze the aeration scheme of the aeration air volume in the aerobic tank based on the parameters in the actual sewage treatment process, which is beneficial to ensuring the reliability of the sewage treatment results.

[0014] It should be understood that the contents described in the Summary of the Invention are not intended to limit the key or important features of the embodiments of the present application, nor are they intended to limit the scope of the present application. Other features of the present application will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The above and other features, advantages and aspects of the embodiments of the present application will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. In the accompanying drawings, the same or similar reference numerals represent the same or similar elements, wherein: Figure 1A schematic diagram of a smart water system in an embodiment of the present application is shown; Figure 2 A flow chart of an intelligent control method applied to a smart water system in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0016] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0017] In addition, the term "and / or" in this article is only a description of the association relationship between the associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0018] The sewage treatment system generally includes anaerobic tanks, anoxic tanks, aerobic tanks and secondary sedimentation tanks which are arranged in sequence. The external sewage to be treated continuously flows through the anaerobic tanks, anoxic tanks, aerobic tanks and secondary sedimentation tanks in sequence and is continuously discharged after meeting the standards, forming a continuous sewage treatment system with continuous reflection. Among them, the anaerobic tank is a closed reaction tank, which is used to use anaerobic microorganisms to carry out the first step of sewage treatment; the anoxic tank is located after the anaerobic tank, and is mainly used for denitrification reaction. It has an internal reflow inlet for receiving the mixed liquid containing nitrate nitrogen returned from the aerobic tank; the aerobic tank is after the anoxic tank, and the aerobic tank is equipped with an aerator. At the same time, there is an internal reflow outlet to send the mixed liquid of nitrate nitrogen back to the anoxic tank, and an external reflow outlet to send the activated sludge back to the front-end anaerobic tank or anoxic tank. The aerobic tank is generally equipped with a dissolved oxygen sensor for real-time monitoring of the dissolved oxygen concentration. The aerobic tank is also provided with an external reflow inlet; the secondary sedimentation tank is arranged after the aerobic tank to achieve mud and water separation. The secondary sedimentation tank is also provided with an external reflow outlet for returning part of the sludge to the aerobic tank.

[0019] In the previous technology, dissolved oxygen sensors are generally used to feedback and adjust the aeration air volume of the aerator to ensure that the dissolved oxygen concentration in the aerobic tank is dynamically stable at the target value and to ensure the effect of sewage treatment. However, the actual complex sewage treatment situation may affect the mechanism of the aeration air volume of the aerator, and only using feedback adjustment is obviously not intelligent enough.

[0020] Therefore, the present application provides an intelligent water system and an intelligent control method for its application, which can take into account various complex situations in the sewage treatment process and accurately control the aeration scheme of the aerator to ensure that the sewage treatment effect is stable and reliable.

[0021] In a first aspect, an embodiment of the present application discloses a smart water system.

[0022] Figure 1 A schematic diagram of a smart water system in an embodiment of the present application is shown.

[0023] Reference Figure 1 The smart water system includes a water volume detection module, a water temperature detection module, a dissolved oxygen detection module, an internal return flow detection module, an external return flow detection module, an aeration air volume detection module, a water quality index acquisition module and a controller.

[0024] Among them, the water volume detection module is used to detect the water volume in the aerobic pool, the water temperature detection module is used to detect the water temperature in the aerobic pool, the dissolved oxygen detection module is used to detect the dissolved oxygen content in the aerobic pool, the internal reflux detection module is used to detect the internal reflux volume from the aerobic pool to the anoxic pool, the external reflux volume detection module is used to detect the external reflux volume from the secondary sedimentation tank to the aerobic pool, and the aeration air volume detection module is used to detect the aeration air volume of the aerator. The above data are all acquired in real time, and the acquired data are all timestamped. Regarding the above detection modules, they are all based on corresponding common knowledge, and there is no special improvement in this application, so they will not be elaborated.

[0025] The water quality index acquisition module is used to obtain the water quality index of the input sewage. The water quality index includes chemical oxygen demand, biochemical oxygen demand, ammonia nitrogen content, phosphorus content, and pH. In the embodiment of the present application, the water quality index is estimated by the staff to estimate the water quality of the sewage to be treated, and the water quality index is entered into the smart water system. That is, the water quality index acquisition module can be specifically manifested as a human-computer interaction facility, such as a touch screen, for the staff to input the water quality index. The water quality index also carries a timestamp. Generally speaking, if the source of the incoming sewage remains unchanged, The controller of the smart water system is equipped with an intelligent control method, which can intelligently analyze the aeration scheme of the aerator according to the water volume, water temperature, dissolved oxygen content, internal return flow, external return flow, water quality index and aeration air volume of the aerobic pool, and determine whether the obtained aeration scheme can accurately meet the needs of sewage treatment, so as to effectively ensure that the treated sewage meets the standards. The intelligent control method in the controller is introduced based on the disclosure of the second aspect of the embodiment of the present application.

[0026] In a second aspect, an embodiment of the present application discloses an intelligent control method applied to a smart water system.

[0027] Figure 2A flow chart of an intelligent control method applied to a smart water system in an embodiment of the present application is shown.

[0028] Reference Figure 2 , the method specifically comprises: S210: Obtain the water volume, water temperature, dissolved oxygen content, internal recirculation volume, external recirculation volume, water quality index and aeration air volume of the aerobic pool within a preset number of unit time periods before the current moment.

[0029] The specific content of the water quality indicators can refer to the disclosure of the first aspect of the embodiment of the present application. The process of acquiring water volume, water temperature, dissolved oxygen content, internal recirculation volume, external recirculation volume, and water quality indicators is a real-time data access process. The water volume, water temperature, and dissolved oxygen content of each unit time are all average values ​​within the unit time. The aeration air volume, internal recirculation volume, and external recirculation volume of each unit time are the cumulative values ​​within the unit time. The values ​​of chemical oxygen demand, biochemical oxygen demand, ammonia nitrogen content, phosphorus content, and pH of each water quality indicator in each unit time are the average values ​​of each water quality indicator, wherein the pH value is specially processed in advance by the pH value, and the neutral value of pH=7 is defined as the zero point. The pH value is equal to the actual pH value minus 7. Acidity is a negative value and alkalinity is a positive value. When calculating the average value, the treated pH is used for calculation.

[0030] S220: Input the water volume, water temperature, dissolved oxygen content, internal recirculation volume, external recirculation volume, water quality index and aeration air volume of the aerobic pool within a preset number of unit time periods before the current moment into a pre-constructed aeration scheme analysis model to obtain the adjustment amount of the aeration air volume and the application time of the adjusted aeration air volume.

[0031] The aeration scheme analysis model is a neural network model, which includes an input layer, multiple hidden layers and an output layer. The input layer includes input nodes corresponding to the water volume, water temperature, dissolved oxygen content, internal recirculation volume, external recirculation volume, water quality indicators and aeration air volume of the aerobic pool within a preset number of unit time periods before the current moment. The output layer includes two output nodes for outputting the adjustment amount of the aeration air volume and the application time of the adjusted aeration air volume.

[0032] In a specific example, the preset number is n, and the water volume, water temperature, dissolved oxygen, internal recirculation, external recirculation, chemical oxygen demand, biochemical oxygen demand, ammonia nitrogen content, phosphorus content, and pH value within the preset number of unit time lengths total 10n data. These 10n data are used to input the input nodes of the aeration scheme analysis model, that is, the total number of input nodes of the input layer is 10n. There are three hidden layers, the first hidden layer has 20 neurons, that is, , which is used for conventional feature extraction and preliminary information fusion. The content of the jth neuron in the first hidden layer is ,in, is an activation function, such as a sigmoid function or a ReLU function, represents the data of the i-th input node, represents the bias of the jth neuron in the second layer, is the calculated weight of the i-th input node for the j-th neuron, is the output of the jth neuron.

[0033] The second hidden layer has 15 neurons, , used for further feature extraction and information fusion, each neuron in the second hidden layer fully links the outputs of all neurons in the first hidden layer, and the content of the kth neuron in the second hidden layer is ,in, is the activation function, is the output of the jth neuron in the first layer, is the calculation weight of the jth neuron in the first layer for the kth neuron in the second layer, is the bias of the kth neuron in the second layer, is the output of the kth neuron in the second layer.

[0034] The third hidden layer has 10 neurons, i.e. , used for further feature extraction and information fusion, each neuron in the third hidden layer fully links the outputs of all neurons in the second hidden layer, and the content of the lth neuron in the third hidden layer is ,in, is the activation function, is the output of the kth neuron in the second layer, is the calculation weight of the kth neuron in the second layer to the lth neuron in the third layer, is the bias of the lth neuron in the third layer, is the output of the lth neuron in the third layer.

[0035] The output layer has two output nodes, and the contents of the two output nodes are and ,in, is the adjustment amount of aeration air volume, is the duration of the adjusted aeration air volume. In the previous formula, is the lth neuron pair in the third layer The calculation weight of is the lth neuron pair in the third layer The calculation weight of , They are , The offset amount.

[0036] For the above-mentioned aeration scheme analysis model, some conventional processing methods of neural network models can also be used, such as data preprocessing to normalize the values ​​of input nodes and map them to For the training method of this model, a multi-objective loss function can be used for training, for example: , in, It is based on the loss of dissolved oxygen in the aerobic pool and the deviation from the target range, such as: , in, is the number of training samples, It is the predicted dissolved oxygen content of the aerobic pool based on the input sample (calculated by adding the current aeration air volume to the predicted aeration air volume adjustment). and It is the target range of dissolved oxygen in the aerobic pool.

[0037] It is the mean square error loss based on the deviation between the predicted aeration air volume and the actual required air volume: , in, It is The actual aeration air volume adjustment of each sample, The model predicts Aeration air volume adjustment for each sample.

[0038] is the mean square error loss based on the deviation between the predicted aeration time and the actual required time: , in, is the actual aeration time of the sth sample, is the aeration time of the sth sample predicted by the model. and is a trade-off coefficient used to adjust the importance of different loss terms.

[0039] Regarding the back propagation and optimization process of the model, you can refer to the following: Use stochastic gradient descent (SGD) or other optimization algorithms with adaptive learning rates (such as Adam) to update weights and biases. Use the Adam optimizer formula described above to update weights: , By using the above aeration scheme analysis model, the adjustment amount of aeration air volume and the duration of the adjusted aeration air volume can be obtained, which is conducive to the precise control of aeration air volume. And the model is continuously trained and optimized during use, so that the model can achieve more accurate aeration scheme control.

[0040] In another example, the aeration scheme analysis model may be set to include four hidden layers, with the first layer having 15 neurons, the second layer including 10 neurons, the third layer including 8 neurons, and the fourth layer including 5 neurons.

[0041] The aeration scheme analysis model is further specially designed so that the aeration scheme analysis model also includes a fluctuation value calculation submodel and a change rate calculation submodel. The fluctuation value calculation submodel and the change rate calculation submodel are both connected to the input nodes of the dissolved oxygen content of a preset number of unit time lengths in the input layer, so that the fluctuation value calculation submodel outputs the dissolved oxygen content fluctuation value, and the change rate calculation submodel outputs the dissolved oxygen content change rate. The dissolved oxygen content fluctuation value is positively correlated to the standard deviation of the dissolved oxygen content of the aerobic pool within the preset number of unit time lengths and / or the absolute value of the difference between the dissolved oxygen content of the aerobic pools within any two adjacent unit time lengths, and the dissolved oxygen content change rate is positively correlated to the result of subtracting the dissolved oxygen content of the aerobic pools in the latter of the two adjacent unit time lengths from the dissolved oxygen content of the former; The hidden layer includes a first special neuron, and the first special neuron is used to input the output of each neuron in the previous hidden layer and the dissolved oxygen fluctuation value and the dissolved oxygen change rate.

[0042] Specifically, the fluctuation value calculation sub-model includes: Assume that the preset number is n, and the dissolved oxygen content of the i-th unit time is , then the average value of dissolved oxygen , standard deviation of dissolved oxygen ; The sum of the absolute values ​​of the differences in dissolved oxygen at adjacent time steps ; The dissolved oxygen fluctuation value , where and are the pre-acquired standard deviation fluctuation coefficient and step difference fluctuation coefficient respectively.

[0043] The change rate calculation sub-model includes: Assume that the preset number is n, and the dissolved oxygen content of the i-th unit time is , calculate the average value of the difference between adjacent time steps of dissolved oxygen , the rate of change of dissolved oxygen is equal to the average value of the difference between adjacent time steps of dissolved oxygen .

[0044] Based on the above, the first special neuron is specifically set in the second layer, and the specific content of the first special neuron is: , in, is the activation function, is the calculation weight of the dissolved oxygen fluctuation value for the first special neuron of the second layer, is the calculation weight of the dissolved oxygen change rate for the kth first special neuron in the second layer.

[0045] In another example, the aeration scheme analysis model may be further designed, wherein the aeration scheme analysis model further includes a water quality fluctuation analysis model and a fluctuation trend analysis model; The water quality fluctuation analysis model and the fluctuation trend analysis model are both connected to the input nodes of the water quality indicators of a preset number of unit time lengths in the input layer, so that the water quality fluctuation analysis sub-model outputs water quality fluctuation data, and the fluctuation trend analysis sub-model outputs fluctuation trend data, the water quality fluctuation data is positively correlated to the standard deviation of the chemical oxygen demand, and / or biochemical oxygen demand, and / or ammonia nitrogen content, and / or phosphorus content, and / or pH within the preset number of unit time lengths, and the fluctuation trend data is positively correlated to the change rate of the chemical oxygen demand, and / or biochemical oxygen demand, and / or ammonia nitrogen content, and / or phosphorus content, and / or pH within the preset number of unit time lengths; The hidden layer includes a second special neuron, which is used to input the output of each neuron in the previous hidden layer and the water quality fluctuation data and fluctuation trend data.

[0046] The water quality fluctuation analysis sub-model includes: Assuming the preset number is n, the chemical oxygen demand, biochemical oxygen demand, ammonia nitrogen content, phosphorus content, and pH value of the i-th unit time are respectively , , , , , calculate the average values ​​of chemical oxygen demand, biochemical oxygen demand, ammonia nitrogen content, phosphorus content, and pH, respectively. , , , , , Then calculate the standard deviation of chemical oxygen demand, biochemical oxygen demand, ammonia nitrogen content, phosphorus content, and pH. , , , , , After calculating the fluctuation trend data .

[0047] The fluctuation trend analysis sub-model includes: Assuming the preset number is n, the chemical oxygen demand, biochemical oxygen demand, ammonia nitrogen content, phosphorus content, and pH value of the i-th unit time are respectively , , , , ; Linear regression model was used for chemical oxygen demand, biochemical oxygen demand, ammonia nitrogen content, phosphorus content, and pH and the least squares method to fit a straight line, where y is the chemical oxygen demand, or the biochemical oxygen demand, or the ammonia nitrogen content, or the phosphorus content, or the pH value, and x is the time in which the unit time is located; The fitting lines of chemical oxygen demand, biochemical oxygen demand, ammonia nitrogen content, phosphorus content, and pH are, , , , , , Determine the slopes of the fitted lines for chemical oxygen demand, biochemical oxygen demand, ammonia nitrogen content, phosphorus content, and pH, and calculate the fluctuation trend data based on the slopes. , In the formula, They are chemical oxygen demand, biochemical oxygen demand, ammonia nitrogen content, phosphorus content, and pH fluctuation trend attention coefficient.

[0048] Based on the above content, the second special neuron is specifically set in the second layer, and the specific content of the second special neuron is: , In the formula, is the activation function, is the calculation weight of the water quality fluctuation data on the second special neuron in the second layer, is the calculation weight of the fluctuation trend data for the second special neuron in the second layer.

[0049] With the aforementioned supplementation, the model can cope with the fluctuation of dissolved oxygen in the aerobic pool, the temporal changes, and the fluctuation and trend changes of water quality indicators to make better predictions, which makes the model have better performance and is conducive to achieving more precise control of the aeration scheme.

[0050] It should be noted that, for the aforementioned method embodiments, for the sake of simplicity of description, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the present application is not limited by the described order of actions, because according to the embodiments of the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily required by the present application.

[0051] To sum up, the present application includes at least the following beneficial effects: it provides an intelligent water system and an intelligent control method for its application, which can intelligently analyze the aeration scheme of the aeration air volume in the aerobic tank based on the parameters in the actual sewage treatment process, which is conducive to ensuring the reliability of the sewage treatment results.

[0052] The above description is only a preferred embodiment of the present application and an explanation of the technical principles used. Those skilled in the art should understand that the scope of disclosure involved in the present application is not limited to the technical solution formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the aforementioned disclosed concept. For example, the above features are replaced with the technical features with similar functions disclosed in this application (but not limited to) by each other to form a technical solution.

Claims

1. An intelligent control method applied to a smart water system, characterized in that: include: Obtain the water volume, water temperature, dissolved oxygen content, internal return flow, external return flow, water quality indicators and aeration air volume of the aerobic pool within a preset number of unit time before the current time, wherein the water quality indicators include chemical oxygen demand, biochemical oxygen demand, ammonia nitrogen content, phosphorus content, and pH; The water volume, water temperature, dissolved oxygen content, internal recirculation volume, external recirculation volume, water quality index and aeration air volume of the aerobic pool within a preset number of unit time periods before the current moment are input into a pre-constructed aeration scheme analysis model to obtain the adjustment amount of the aeration air volume and the application time of the adjusted aeration air volume. The aeration scheme analysis model includes an input layer, multiple hidden layers and an output layer. The input layer includes input nodes that correspond one to one with the water volume, water temperature, dissolved oxygen content, internal recirculation volume, external recirculation volume, water quality index and aeration air volume of the aerobic pool within a preset number of unit time periods before the current moment. The output layer includes two output nodes for outputting the adjustment amount of the aeration air volume and the application time of the adjusted aeration air volume.

2. The method according to claim 1, characterized in that The aeration scheme analysis model also includes a fluctuation value calculation submodel and a change rate calculation submodel; The fluctuation value calculation submodel and the change rate calculation submodel are both connected to the input nodes of the dissolved oxygen of a preset number of unit time lengths in the input layer, so that the fluctuation value calculation submodel outputs the dissolved oxygen fluctuation value, and the change rate calculation submodel outputs the dissolved oxygen change rate, the dissolved oxygen fluctuation value is positively correlated to the standard deviation of the dissolved oxygen in the aerobic pool within the preset number of unit time lengths and / or the absolute value of the difference between the dissolved oxygen in any two adjacent unit time lengths of the aerobic pool, and the dissolved oxygen change rate is positively correlated to the result of subtracting the dissolved oxygen in the aerobic pool in the latter of the two adjacent unit time lengths from the dissolved oxygen in the former; The hidden layer includes a first special neuron, and the first special neuron is used to input the output of each neuron in the previous hidden layer and the dissolved oxygen fluctuation value and the dissolved oxygen change rate.

3. The method according to claim 2, characterized in that The fluctuation value calculation sub-model includes: Assume that the preset number is n, and the dissolved oxygen content of the i-th unit time is , then the average value of dissolved oxygen , standard deviation of dissolved oxygen ; The sum of the absolute values ​​of the differences in dissolved oxygen at adjacent time steps ; The dissolved oxygen fluctuation value , where and are the pre-acquired standard deviation fluctuation coefficient and step difference fluctuation coefficient respectively.

4. The method according to claim 2, characterized in that: The change rate calculation sub-model includes: Assume that the preset number is n, and the dissolved oxygen content of the i-th unit time is , calculate the average value of the difference between adjacent time steps of dissolved oxygen , the rate of change of dissolved oxygen is equal to the average value of the difference between adjacent time steps of dissolved oxygen .

5. The method according to claim 1, characterized in that The aeration scheme analysis model also includes a water quality fluctuation analysis model and a fluctuation trend analysis model; The water quality fluctuation analysis model and the fluctuation trend analysis model are both connected to the input nodes of the water quality indicators of a preset number of unit time lengths in the input layer, so that the water quality fluctuation analysis sub-model outputs water quality fluctuation data, and the fluctuation trend analysis sub-model outputs fluctuation trend data, the water quality fluctuation data is positively correlated to the standard deviation of the chemical oxygen demand, and / or biochemical oxygen demand, and / or ammonia nitrogen content, and / or phosphorus content, and / or pH within the preset number of unit time lengths, and the fluctuation trend data is positively correlated to the change rate of the chemical oxygen demand, and / or biochemical oxygen demand, and / or ammonia nitrogen content, and / or phosphorus content, and / or pH within the preset number of unit time lengths; The hidden layer includes a second special neuron, which is used to input the output of each neuron in the previous hidden layer and the water quality fluctuation data and fluctuation trend data.

6. The method according to claim 5, characterized in that The water quality fluctuation analysis sub-model includes: Assuming the preset number is n, the chemical oxygen demand, biochemical oxygen demand, ammonia nitrogen content, phosphorus content, and pH value of the i-th unit time are respectively , , , , , calculate the average values ​​of chemical oxygen demand, biochemical oxygen demand, ammonia nitrogen content, phosphorus content, and pH, respectively. ， ， ， ， ， Then calculate the standard deviation of chemical oxygen demand, biochemical oxygen demand, ammonia nitrogen content, phosphorus content, and pH. ， ， ， ， ， After calculating the fluctuation trend data .

7. The method according to claim 5, characterized in that The fluctuation trend analysis sub-model includes: Assuming the preset number is n, the chemical oxygen demand, biochemical oxygen demand, ammonia nitrogen content, phosphorus content, and pH value of the i-th unit time are respectively , , , , ; Linear regression model was used for chemical oxygen demand, biochemical oxygen demand, ammonia nitrogen content, phosphorus content, and pH and the least squares method to fit a straight line, where y is the chemical oxygen demand, or the biochemical oxygen demand, or the ammonia nitrogen content, or the phosphorus content, or the pH value, and x is the time per unit time; The fitting lines of chemical oxygen demand, biochemical oxygen demand, ammonia nitrogen content, phosphorus content, and pH are, respectively, , ， ， ， ， Determine the slopes of the fitted lines for chemical oxygen demand, biochemical oxygen demand, ammonia nitrogen content, phosphorus content, and pH, and calculate the fluctuation trend data based on the slopes. ， In the formula, They are chemical oxygen demand, biochemical oxygen demand, ammonia nitrogen content, phosphorus content, and pH fluctuation trend attention coefficient.

8. A smart water system, characterized in that: It comprises a controller, which is used to control the adjustment amount of the aeration air volume in the aerobic pool and the application time of the adjusted aeration air volume, and the controller is configured to execute the method according to any one of claims 1-7.

Images