Sewage system regulation and control method based on intelligent water quality prediction

Abstract

The invention discloses a sewage system regulation and control method based on intelligent water quality prediction, and relates to the field of reflux ratio regulation and control in water treatment.The method comprises the steps that auxiliary variables are determined, and an anoxic pond effluent biochemical oxygen demand prediction model or an anoxic pond effluent nitrate concentration prediction model is constructed; regulating and controlling the reflux ratio, and carrying out delay checking. When a water quality instrument has no signal, the method can temporarily predict the real effluent quality data of the anoxic tank, participates in refinement regulation and control of the reflux ratio, ensures that the system can continuously and stably operate for more than 3 months when the instrument is unstable, and the deviation degree of the predicted mean value is less than 10% compared with the real value. By predicting the effluent quality data of the anoxic zone, the correlation logic of the reflux ratio and the related water quality data is formed, and the reflux ratio is finely regulated, so that the condition that the total nitrogen removal rate is limited due to the too low reflux ratio or the chemical oxygen demand of the anoxic zone penetrates into the aerobic zone to influence nitrification is avoided; the influence on an anoxic environment caused by too high dissolved oxygen due to too high reflux ratio is also avoided.

Description

Technical Field

[0001] The present invention relates to the field of reflow ratio control in water treatment, and in particular to a sewage system control method based on intelligent water quality prediction. Background Art

[0002] Sewage treatment is based on the AAO process. The regulation of the reflow ratio is related to the compliance of the effluent TN and the cost of the carbon source, and is an important parameter for operation and regulation. The effluent quality of the anoxic tank is an important basis for adjusting the reflow ratio, which requires additional instrumentation and probe support, increasing the investment in instrumentation and equipment in the sewage plant. With the development of technologies such as artificial intelligence and soft measurement, the use of intelligent prediction to predict the effluent quality of the anoxic zone and adjust the reflow ratio in real time is an effective means to solve the existing problems. In the prior art, methods have been proposed to solve this problem, but they all have certain limitations.

[0003] In terms of water quality prediction: CN107664682A discloses a water quality soft measurement prediction method for ammonia nitrogen. This method forms a data sample set by acquiring the data values ​​of multiple water quality monitoring indicators including ammonia nitrogen in the water environment to be tested, and further divides it into a training set and a test set. The training set is trained using a fuzzy neural network algorithm to obtain an ammonia nitrogen soft measurement model; the model is tested using the test set, and the model establishment and testing are repeatedly performed until the test results obtained meet the preset conditions, and the ammonia nitrogen soft measurement model that meets the preset conditions is used as the final soft measurement model.

[0004] CN118298962A discloses a method for predicting influent ammonia nitrogen concentration based on long short-term memory step sequence (LSTM). The method first constructs prediction input data and output data, and preprocesses the data, then uses the LSTM model to predict the ammonia nitrogen concentration, sets the model parameters to improve the prediction accuracy, and finally simulates the ammonia nitrogen concentration of the sewage plant influent and evaluates the prediction performance of the model.

[0005] CN117711521A discloses a method for predicting the ammonia nitrogen content of compost using an artificial intelligence model. The method obtains environmental parameters and ammonia nitrogen content information during the composting process of organic solid waste, inputs this information into an elastic network to select the optimal environmental parameter combination, and then inputs it into an artificial neural network to predict the ammonia nitrogen content.

[0006] CN118886529A provides a method for predicting water quality in a sewage treatment plant based on an improved LSTM neural network model. The method first obtains water quality monitoring data from the sewage treatment plant, performs outlier cleaning, smoothing and normalization on the data, and then optimizes the LSTM model using a genetic algorithm, performs training and testing to obtain a water quality prediction model.

[0007] The problem with this type of patent is:

[0008] ① The amount of data required is large and the investment cost is high. The ammonia nitrogen prediction model is completely dependent on neural network training and requires massive data support, which increases the requirements for operating data; CN107664682A mentioned that in the implementation method, 24 water quality indicators were selected for modeling through three modelings; CN 117711521 A mentioned that the environmental information to be predicted in the composting process of organic solid waste includes many parameters such as number of days, temperature, total carbon, total nitrogen, compost volume, pH value, moisture content, carbon-nitrogen ratio, seed germination rate, conductivity, etc.; the influencing factors of ammonia nitrogen concentration in CN118298962A include PH, suspended solids (SS), water temperature (WT), air temperature, air pressure, relative humidity, rainfall, wind speed, average total cloud cover and visibility, etc. Too many indicators are required, which requires the support of instrument probes, and thus increases investment costs;

[0009] ② Low data utilization. CN107664682A optimizes the model through repeated training and testing, which increases development costs and may also lead to a long model training cycle, affecting the speed and efficiency of practical applications. Excessive data demand also complicates data collection and processing, reducing the response speed and flexibility of the system.

[0010] ③ Poor prediction accuracy. Its input data is completely based on instrument measurement data, and it is impossible to rule out the confusion of correlation caused by poor instrument measurement accuracy, which in turn affects the accuracy of the model. In addition, a single model is difficult to handle complex and changeable water quality conditions, especially when facing different types of sewage, which may lead to a decrease in prediction accuracy.

[0011] ④ Lack of feedback mechanism. Most existing technologies lack an effective feedback mechanism and are unable to self-correct and optimize based on the difference between actual operating data and predicted results. This defect causes the model to gradually lose accuracy in long-term operation, especially when faced with sudden changes in water quality, and the prediction strategy cannot be adjusted in time. Although CN107664682A and CN118298962A have conducted multiple model training and testing, they did not introduce a feedback loop of actual operating data, making it difficult to achieve continuous improvement.

[0012] In terms of reflux ratio control: CN115448531B discloses a method for correcting the internal and external reflux ratio of the A2 / O process. First, the external reflux ratio is calculated based on the known data of the sewage treatment line, and the initial sampling period is calculated in combination with the initially given internal reflux ratio. Then, the sludge in the anoxic tank is sampled at the same intervals during the sampling period to obtain the change in phosphate concentration of the anoxic tank sludge during the sampling period. At the same time, the phosphate concentration of the effluent from the anaerobic tank, the phosphate concentration of the internal reflux sewage, and the phosphate concentration of the effluent from the anoxic tank are detected respectively to calculate the actual internal reflux ratio. The sampling period is recalibrated using the actual internal reflux ratio, and the sampling is re-measured to further calibrate the internal reflux ratio. The iteration is continued until the coefficient of variation of the internal reflux ratio measured by two samplings is less than 0.1, and the internal reflux ratio correction is completed. This patent has the following defects: first, the method relies on actual sampling and detection data, the calibration process is time-consuming, and it is difficult to meet the needs of real-time regulation; second, the sampling and detection process may be affected by various factors, resulting in limited accuracy of the calibration results; third, the method lacks effective response strategies when the instrument fails or the data is abnormal, and the system stability and robustness are insufficient.

[0013] In summary, although the existing technology has achieved monitoring and prediction functions to a certain extent in the water quality concentration prediction method, there are still problems such as large data demand, low data utilization, poor prediction accuracy and no feedback mechanism in practical applications. In the control of the reflow ratio, it relies on actual sampling and detection data, the correction process is time-consuming, and may be affected by various factors, resulting in limited accuracy of the correction results. Therefore, the next direction of improvement should focus on developing an intelligent control system that can accurately predict the effluent quality of the anoxic tank, adjust the reflow ratio in real time, and have high stability and robustness. Such a system should be able to make full use of existing data, improve data utilization, reduce investment costs, and be able to quickly adjust the prediction and control strategies in the face of sudden changes in water quality to ensure stable compliance of effluent water quality and effective control of operating costs. Summary of the invention

[0014] In view of the problems existing in the prior art, the present invention provides a sewage system control method based on intelligent water quality prediction.

[0015] The present invention adopts the following technical solutions:

[0016] A sewage system control method based on water quality intelligent prediction includes the following steps:

[0017] Step 1: Determine the auxiliary variables; if there is no signal from the biochemical oxygen demand instrument of the anoxic pool effluent, determine the first auxiliary variable as the nitrate nitrogen concentration of the anoxic pool effluent, the anoxic pool reflow ratio, the redox potential, the pH value and the water temperature, and obtain the historical data of the biochemical oxygen demand of the anoxic pool effluent and the first auxiliary variable;

[0018] If there is no signal from the nitrate nitrogen meter in the anoxic pool outlet water, the second auxiliary variable is determined to be the biochemical oxygen demand of the anoxic pool outlet water, the anoxic pool reflow ratio, the redox potential, the pH value and the water temperature, and the historical data of the nitrate nitrogen concentration in the anoxic pool outlet water and the second auxiliary variable are obtained;

[0019] Step 2: construct a model; construct a functional relationship between the biochemical oxygen demand of the effluent from the anoxic pool and the first auxiliary variable according to the historical data of the biochemical oxygen demand of the effluent from the anoxic pool and the first auxiliary variable, and form a biochemical oxygen demand prediction model for the effluent from the anoxic pool; input the first auxiliary variable at the time to be predicted into the biochemical oxygen demand prediction model for the effluent from the anoxic pool, and obtain the biochemical oxygen demand of the effluent from the anoxic pool at the time to be predicted as the prediction data;

[0020] According to the historical data of the nitrate-nitrogen concentration in the effluent of the anoxic pool and the second auxiliary variable, a functional relationship between the nitrate-nitrogen concentration in the effluent of the anoxic pool and the second auxiliary variable is constructed to form a prediction model for the nitrate-nitrogen concentration in the effluent of the anoxic pool; the second auxiliary variable at the time to be predicted is input into the prediction model for the nitrate-nitrogen concentration in the effluent of the anoxic pool to obtain the prediction data for the nitrate-nitrogen concentration in the effluent of the anoxic pool at the time to be predicted;

[0021] Step 3: Regulate the reflow ratio; if the nitrate nitrogen concentration of the anoxic pool effluent is less than 1 mg / L and the biochemical oxygen demand of the anoxic pool effluent is less than 50 mg / L, the reflow ratio is maintained unchanged;

[0022] If the nitrate nitrogen concentration of the effluent from the anoxic pool is less than 1 mg / L and the biochemical oxygen demand of the effluent from the anoxic pool is ≥50 mg / L, the reflow ratio will be increased by 10%;

[0023] If the nitrate nitrogen concentration in the anoxic pool effluent is ≥1 mg / L, the reflux ratio will be reduced by 10%;

[0024] Step 4: Delayed calibration; after the prediction data is run, 25%-50% of the actual hydraulic retention time of the anoxic tank is used as the calibration interval, the value of the prediction data corresponding to 25% of the actual hydraulic retention time of the anoxic tank is C1, and the value of the prediction data corresponding to 50% of the actual hydraulic retention time of the anoxic tank is C2;

[0025] When the predicted data is the nitrate-nitrogen concentration value of the effluent from the anoxic pool, if |C1-C2|≤10%C1, the nitrate-nitrogen concentration value of the effluent from the anoxic pool remains unchanged; when C2-C1>10%C1, the compensation coefficient k is set to 1.2, and the final predicted value is 1.2*the nitrate-nitrogen concentration value of the effluent from the anoxic pool; when C1-C2>10%C1, the compensation coefficient k is set to 0.8, and the final predicted value is 0.8*the nitrate-nitrogen concentration value of the effluent from the anoxic pool;

[0026] When the predicted data is the biochemical oxygen demand of the effluent from the anoxic pool, if |C1-C2|≤10%C1, the biochemical oxygen demand of the effluent from the anoxic pool remains unchanged; when C2-C1>10%C1, the compensation coefficient k is set to 0.8, and the final predicted value is 0.8*biochemical oxygen demand of the effluent from the anoxic pool; when C1-C2>10%C1, the compensation coefficient k is set to 1.2, and the final predicted value is 1.2*biochemical oxygen demand of the effluent from the anoxic pool.

[0027] Preferably, the functional relationship between the biochemical oxygen demand of the effluent from the anoxic pool and the first auxiliary variable is constructed by a multi-form regression algorithm, and the functional relationship between the nitrate-nitrogen concentration of the effluent from the anoxic pool and the second auxiliary variable is constructed by a multi-form regression algorithm.

[0028] The present invention has the following beneficial effects:

[0029] 1) System stability. When there is no signal from the water quality meter, the water quality data of the anoxic pool outlet can be predicted temporarily, and the reflow ratio can be finely controlled to ensure that the system can continue to operate stably for more than 3 months when the meter is unstable, and the deviation of the predicted mean from the actual value is less than 10%.

[0030] 2) High effluent quality. Through the prediction of effluent quality data in the anoxic zone, the correlation logic between the reflow ratio and related water quality data is formed, and then the reflow ratio is finely regulated to avoid the limited total nitrogen removal rate caused by too low reflow ratio or the penetration of chemical oxygen demand in the anoxic zone into the aerobic zone to affect nitrification, and to avoid the high effluent ratio causing too high dissolved oxygen to affect the anoxic environment. The total nitrogen in the effluent can be stably lower than 5 mg / L, meeting the current optimal TN emission standard.

[0031] 3) Save medicine consumption. Through the prediction of effluent water quality data in the anoxic zone and the adjustment of the reflow ratio, the utilization of raw water carbon source is strengthened, the amount of external carbon source is reduced by more than 50%, and the medicine consumption is reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 This is the total nitrogen diagram of the inlet and outlet water of Example 1. DETAILED DESCRIPTION

[0033] The specific implementation of the present invention is further described below in conjunction with the accompanying drawings and specific embodiments:

[0034] Example 1

[0035] Combination Figure 1 , a sewage treatment plant, the effluent total nitrogen standard is 5mg / L, and the biochemical oxygen demand instrument of the anoxic tank effluent of the biochemical section of the water plant has no signal at the predicted time.

[0036] Step 1: Taking the nitrate nitrogen concentration of the anoxic pool effluent, the reflow ratio of the anoxic pool, the redox potential, the pH value and the water temperature as the first auxiliary variable, obtain the 60-day historical data of the biochemical oxygen demand of the anoxic pool effluent and the first auxiliary variable.

[0037] Step 2: Based on the 60-day historical data of the biochemical oxygen demand of the anoxic pool effluent and the first auxiliary variable, a multi-form regression algorithm is used to construct a functional relationship between the biochemical oxygen demand of the anoxic pool effluent and the first auxiliary variable to form an anoxic pool effluent biochemical oxygen demand prediction model. The first auxiliary variable at the time to be predicted is input into the anoxic pool effluent biochemical oxygen demand prediction model, and the anoxic pool effluent biochemical oxygen demand at the time to be predicted is obtained as the prediction data.

[0038] Step 3: If the nitrate-nitrogen concentration of the effluent from the anoxic pool is less than 1 mg / L, and the biochemical oxygen demand of the effluent from the anoxic pool at the predicted time is less than 50 mg / L, the reflow ratio is maintained unchanged.

[0039] If the nitrate-nitrogen concentration of the effluent from the anoxic pool is less than 1 mg / L, and the biochemical oxygen demand of the effluent from the anoxic pool at the predicted time is ≥50 mg / L, the reflow ratio will be increased by 10%.

[0040] If the nitrate-nitrogen concentration in the effluent from the anoxic pool is ≥1 mg / L, the reflux ratio will be reduced by 10%.

[0041] Step 4: After running the predicted data, 25%-50% of the actual hydraulic retention time of the anoxic pool is used as the calibration interval. The value of the biochemical oxygen demand of the anoxic pool effluent corresponding to 25% of the actual hydraulic retention time of the anoxic pool is C1, and the value of the biochemical oxygen demand of the anoxic pool effluent corresponding to 50% of the actual hydraulic retention time of the anoxic pool is C2.

[0042] If |C1-C2|≤10%C1, the biochemical oxygen demand of the effluent from the anoxic pool remains unchanged; when C2-C1>10%C1, the compensation coefficient k is set to 0.8, and the final predicted value is 0.8*biochemical oxygen demand of the effluent from the anoxic pool; when C1-C2>10%C1, the compensation coefficient k is set to 1.2, and the final predicted value is 1.2*biochemical oxygen demand of the effluent from the anoxic pool.

[0043] like Figure 1 As shown, the operating data of the water plant after more than 400 days of operation showed that the total nitrogen in the effluent using the method of the present invention was stabilized at 3.80±0.81 mg / L, with a compliance rate of 100%, and the C / N of the denitrification consumption was calculated to be 4.46±0.78.

[0044] Example 2

[0045] The total nitrogen standard of the effluent of a sewage treatment plant is 15 mg / L. At the predicted time, there is no signal from the nitrate nitrogen meter in the effluent of the anoxic tank of the biochemical section of the water plant.

[0046] Step 1: Take the biochemical oxygen demand of the effluent from the anoxic pool, the reflow ratio of the anoxic pool, the redox potential, the pH value and the water temperature as the second auxiliary variable, and obtain the 60-day historical data of the effluent nitrate nitrogen concentration and the second auxiliary variable of the anoxic pool.

[0047] Step 2: Based on the 60-day historical data of the nitrate-nitrogen concentration of the anoxic pool effluent and the second auxiliary variable, a prediction model for the nitrate-nitrogen concentration of the anoxic pool effluent is formed by using the multi-form regression algorithm to calculate the functional relationship between the nitrate-nitrogen concentration of the anoxic pool effluent and the second auxiliary variable. The second auxiliary variable at the time to be predicted is input into the prediction model for the nitrate-nitrogen concentration of the anoxic pool effluent to obtain the predicted data for the nitrate-nitrogen concentration of the anoxic pool effluent at the time to be predicted.

[0048] Step 3: If the nitrate-nitrogen concentration of the effluent from the anoxic pool at the time to be predicted is less than 1 mg / L, and the biochemical oxygen demand of the effluent from the anoxic pool is less than 50 mg / L, the reflow ratio is maintained unchanged.

[0049] If the nitrate-nitrogen concentration of the effluent from the anoxic pool at the time of prediction is less than 1 mg / L and the biochemical oxygen demand of the effluent from the anoxic pool is ≥50 mg / L, the reflux ratio will be increased by 10%.

[0050] If the nitrate-nitrogen concentration of the anoxic pool effluent at the predicted time is ≥1 mg / L, the reflux ratio will be reduced by 10%.

[0051] Step 4: After running the predicted data, 25%-50% of the actual hydraulic retention time of the anoxic pool is used as the calibration interval. The value of the nitrate-nitrogen concentration in the effluent of the anoxic pool corresponding to 25% of the actual hydraulic retention time of the anoxic pool is C1, and the value of the nitrate-nitrogen concentration in the effluent of the anoxic pool corresponding to 50% of the actual hydraulic retention time of the anoxic pool is C2.

[0052] If |C1-C2|≤10%C1, the nitrate-nitrogen concentration in the effluent from the anoxic pool remains unchanged; when C2-C1>10%C1, the compensation coefficient k is set to 1.2, and the final predicted value is 1.2*the nitrate-nitrogen concentration in the effluent from the anoxic pool; when C1-C2>10%C1, the compensation coefficient k is set to 0.8, and the final predicted value is 0.8*the nitrate-nitrogen concentration in the effluent from the anoxic pool.

[0053] The 200-day operation data of the water plant showed that the total nitrogen in the effluent using the method of the present invention was 12.69±0.85 mg / L, with a compliance rate of 100%, and the C / N required for denitrification was 4.89±1.26.

[0054] Comparative Example 1

[0055] In a sewage pilot system, the total nitrogen standard of the effluent is 15 mg / L, and there is no signal from the biochemical oxygen demand instrument of the effluent from the anoxic tank of the pilot biochemical section.

[0056] Step 1: Taking the nitrate nitrogen concentration of the anoxic pool effluent, the reflow ratio of the anoxic pool, the redox potential, the pH value and the water temperature as the first auxiliary variable, obtain the 60-day historical data of the biochemical oxygen demand of the anoxic pool effluent and the first auxiliary variable.

[0057] Step 2: Based on the 60-day historical data of the biochemical oxygen demand of the anoxic pool effluent and the first auxiliary variable, a multi-form regression algorithm is used to construct a functional relationship between the biochemical oxygen demand of the anoxic pool effluent and the first auxiliary variable to form an anoxic pool effluent biochemical oxygen demand prediction model. The first auxiliary variable at the time to be predicted is input into the anoxic pool effluent biochemical oxygen demand prediction model, and the anoxic pool effluent biochemical oxygen demand at the time to be predicted is obtained as the prediction data.

[0058] Step 3: If the nitrate-nitrogen concentration of the effluent from the anoxic pool is less than 2 mg / L, and the biochemical oxygen demand of the effluent from the anoxic pool at the predicted time is less than 50 mg / L, the reflow ratio is maintained unchanged.

[0059] If the nitrate-nitrogen concentration of the effluent from the anoxic pool is less than 2 mg / L, and the biochemical oxygen demand of the effluent from the anoxic pool at the predicted time is ≥50 mg / L, the reflow ratio will be increased by 10%.

[0060] If the nitrate-nitrogen concentration in the effluent from the anoxic tank is ≥2 mg / L, the reflux ratio will be reduced by 10%.

[0061] Step 4: After running the predicted data, 25%-50% of the actual hydraulic retention time of the anoxic pool is used as the calibration interval. The value of the biochemical oxygen demand of the anoxic pool effluent corresponding to 25% of the actual hydraulic retention time of the anoxic pool is C1, and the value of the biochemical oxygen demand of the anoxic pool effluent corresponding to 50% of the actual hydraulic retention time of the anoxic pool is C2.

[0062] If |C1-C2|≤10%C1, the biochemical oxygen demand of the effluent from the anoxic pool remains unchanged; when C2-C1>10%C1, the compensation coefficient k is set to 0.8, and the final predicted value is 0.8*biochemical oxygen demand of the effluent from the anoxic pool; when C1-C2>10%C1, the compensation coefficient k is set to 1.2, and the final predicted value is 1.2*biochemical oxygen demand of the effluent from the anoxic pool.

[0063] The 200-day operation data of the pilot test showed that the TN of the effluent using the method of Comparative Example 1 was 12.59±0.59 mg / L, the compliance rate was 94.41%, and the C / N required for denitrification was 5.23±1.27.

[0064] Comparative Example 2

[0065] In the sewage pilot system in Comparative Example 1, the effluent total nitrogen standard is 15 mg / L, and the biochemical oxygen demand meter of the effluent from the anoxic tank of the pilot biochemical section has no signal.

[0066] Step 1: Taking the nitrate nitrogen concentration of the anoxic pool effluent, the reflow ratio of the anoxic pool, the redox potential, the pH value and the water temperature as the first auxiliary variable, obtain the 60-day historical data of the biochemical oxygen demand of the anoxic pool effluent and the first auxiliary variable.

[0067] Step 2: Based on the 60-day historical data of the biochemical oxygen demand of the anoxic pool effluent and the first auxiliary variable, a multi-form regression algorithm is used to construct a functional relationship between the biochemical oxygen demand of the anoxic pool effluent and the first auxiliary variable to form an anoxic pool effluent biochemical oxygen demand prediction model. The first auxiliary variable at the time to be predicted is input into the anoxic pool effluent biochemical oxygen demand prediction model, and the anoxic pool effluent biochemical oxygen demand at the time to be predicted is obtained as the prediction data.

[0068] Step 3: If the nitrate-nitrogen concentration of the effluent from the anoxic pool is less than 1 mg / L, and the biochemical oxygen demand of the effluent from the anoxic pool at the predicted time is less than 60 mg / L, the reflow ratio is maintained unchanged.

[0069] If the nitrate-nitrogen concentration of the effluent from the anoxic pool is less than 1 mg / L, and the biochemical oxygen demand of the effluent from the anoxic pool at the predicted time is ≥ 60 mg / L, the reflow ratio will be increased by 10%.

[0070] If the nitrate-nitrogen concentration in the effluent from the anoxic pool is ≥1 mg / L, the reflux ratio will be reduced by 10%.

[0071] Step 4: After running the predicted data, 25%-50% of the actual hydraulic retention time of the anoxic pool is used as the calibration interval. The value of the biochemical oxygen demand of the anoxic pool effluent corresponding to 25% of the actual hydraulic retention time of the anoxic pool is C1, and the value of the biochemical oxygen demand of the anoxic pool effluent corresponding to 50% of the actual hydraulic retention time of the anoxic pool is C2.

[0072] If |C1-C2|≤10%C1, the biochemical oxygen demand of the effluent from the anoxic pool remains unchanged; when C2-C1>10%C1, the compensation coefficient k is set to 0.8, and the final predicted value is 0.8*biochemical oxygen demand of the effluent from the anoxic pool; when C1-C2>10%C1, the compensation coefficient k is set to 1.2, and the final predicted value is 1.2*biochemical oxygen demand of the effluent from the anoxic pool.

[0073] The 200-day operation data of the pilot showed that the TN of the effluent using this method was 12.09±0.46 mg / L, with a compliance rate of 100%, but the C / N required for denitrification was 5.51±0.72.

[0074] Comparative Example 3

[0075] In the sewage pilot system in Comparative Example 1, the effluent total nitrogen standard is 15 mg / L, and the biochemical oxygen demand meter of the effluent from the anoxic tank of the pilot biochemical section has no signal.

[0076] Step 1: Taking the nitrate nitrogen concentration of the anoxic pool effluent, the reflow ratio of the anoxic pool, the redox potential, the pH value and the water temperature as the first auxiliary variable, obtain the 60-day historical data of the biochemical oxygen demand of the anoxic pool effluent and the first auxiliary variable.

[0077] Step 2: Based on the 60-day historical data of the biochemical oxygen demand of the anoxic pool effluent and the first auxiliary variable, a multi-form regression algorithm is used to construct a functional relationship between the biochemical oxygen demand of the anoxic pool effluent and the first auxiliary variable to form an anoxic pool effluent biochemical oxygen demand prediction model. The first auxiliary variable at the time to be predicted is input into the anoxic pool effluent biochemical oxygen demand prediction model, and the anoxic pool effluent biochemical oxygen demand at the time to be predicted is obtained as the prediction data.

[0078] Step 3: If the nitrate-nitrogen concentration of the effluent from the anoxic pool is less than 1 mg / L, and the biochemical oxygen demand of the effluent from the anoxic pool at the predicted time is less than 50 mg / L, the reflow ratio is maintained unchanged.

[0079] If the nitrate-nitrogen concentration of the effluent from the anoxic pool is less than 1 mg / L, and the biochemical oxygen demand of the effluent from the anoxic pool at the predicted time is ≥50 mg / L, the reflow ratio will be increased by 10%.

[0080] If the nitrate-nitrogen concentration in the effluent from the anoxic pool is ≥1 mg / L, the reflux ratio will be reduced by 10%.

[0081] Step 4: After running the predicted data, 10%-25% of the actual hydraulic retention time of the anoxic pool is used as the calibration interval. The value of the biochemical oxygen demand of the anoxic pool effluent corresponding to 10% of the actual hydraulic retention time of the anoxic pool is C1, and the value of the biochemical oxygen demand of the anoxic pool effluent corresponding to 25% of the actual hydraulic retention time of the anoxic pool is C2.

[0082] If |C1-C2|≤10%C1, the biochemical oxygen demand of the effluent from the anoxic pool remains unchanged; when C2-C1>10%C1, the compensation coefficient k is set to 0.8, and the final predicted value is 0.8*biochemical oxygen demand of the effluent from the anoxic pool; when C1-C2>10%C1, the compensation coefficient k is set to 1.2, and the final predicted value is 1.2*biochemical oxygen demand of the effluent from the anoxic pool.

[0083] The 200-day operation data of the pilot project showed that the TN of the effluent using this method was 11.55±0.54 mg / L, with a compliance rate of 100%, but the C / N required for denitrification was 5.34±0.74.

[0084] Comparative Example 4

[0085] In the sewage pilot system in Comparative Example 1, the effluent total nitrogen standard is 15 mg / L, and the biochemical oxygen demand meter of the effluent from the anoxic tank of the pilot biochemical section has no signal.

[0086] Step 1: Taking the nitrate nitrogen concentration of the anoxic pool effluent, the reflow ratio of the anoxic pool, the redox potential, the pH value and the water temperature as the first auxiliary variable, obtain the 60-day historical data of the biochemical oxygen demand of the anoxic pool effluent and the first auxiliary variable.

[0087] Step 2: Based on the 60-day historical data of the biochemical oxygen demand of the anoxic pool effluent and the first auxiliary variable, a multi-form regression algorithm is used to construct a functional relationship between the biochemical oxygen demand of the anoxic pool effluent and the first auxiliary variable to form an anoxic pool effluent biochemical oxygen demand prediction model. The first auxiliary variable at the time to be predicted is input into the anoxic pool effluent biochemical oxygen demand prediction model, and the anoxic pool effluent biochemical oxygen demand at the time to be predicted is obtained as the prediction data.

[0088] Step 3: If the nitrate-nitrogen concentration of the effluent from the anoxic pool is less than 1 mg / L, and the biochemical oxygen demand of the effluent from the anoxic pool at the predicted time is less than 50 mg / L, the reflow ratio is maintained unchanged.

[0089] If the nitrate-nitrogen concentration of the effluent from the anoxic pool is less than 1 mg / L, and the biochemical oxygen demand of the effluent from the anoxic pool at the predicted time is ≥50 mg / L, the reflow ratio will be increased by 10%.

[0090] If the nitrate-nitrogen concentration in the effluent from the anoxic pool is ≥1 mg / L, the reflux ratio will be reduced by 10%.

[0091] Step 4: After running the predicted data, 50%-75% of the actual hydraulic retention time of the anoxic pool is used as the calibration interval. The value of the biochemical oxygen demand of the anoxic pool effluent corresponding to 50% of the actual hydraulic retention time of the anoxic pool is C1, and the value of the biochemical oxygen demand of the anoxic pool effluent corresponding to 75% of the actual hydraulic retention time of the anoxic pool is C2.

[0092] If |C1-C2|≤10%C1, the biochemical oxygen demand of the effluent from the anoxic pool remains unchanged; when C2-C1>10%C1, the compensation coefficient k is set to 0.8, and the final predicted value is 0.8*biochemical oxygen demand of the effluent from the anoxic pool; when C1-C2>10%C1, the compensation coefficient k is set to 1.2, and the final predicted value is 1.2*biochemical oxygen demand of the effluent from the anoxic pool.

[0093] The 200-day operation data of the pilot project showed that the TN of the effluent using this method was 12.07±1.18 mg / L, the compliance rate was 91.26%, and the C / N required for denitrification was 4.66±1.67.

[0094] Of course, the above description is not a limitation of the present invention, and the present invention is not limited to the above examples. Changes, modifications, additions or substitutions made by technicians in this technical field within the essential scope of the present invention should also fall within the protection scope of the present invention.

Claims

1. A sewage system control method based on intelligent water quality prediction, characterized in that: The following steps are involved: Step 1: Determine auxiliary variables; If there is no signal from the biochemical oxygen demand instrument of the anoxic pool effluent, the first auxiliary variable is determined to be the nitrate nitrogen concentration of the anoxic pool effluent, the anoxic pool reflow ratio, the redox potential, the pH value and the water temperature, and the historical data of the biochemical oxygen demand of the anoxic pool effluent and the first auxiliary variable are obtained; If there is no signal from the nitrate nitrogen meter in the anoxic pool outlet water, the second auxiliary variable is determined to be the biochemical oxygen demand of the anoxic pool outlet water, the anoxic pool reflow ratio, the redox potential, the pH value and the water temperature, and the historical data of the nitrate nitrogen concentration in the anoxic pool outlet water and the second auxiliary variable are obtained; Step 2: Build the model; According to the historical data of the biochemical oxygen demand of the effluent from the anoxic pool and the first auxiliary variable, a functional relationship between the biochemical oxygen demand of the effluent from the anoxic pool and the first auxiliary variable is constructed to form a biochemical oxygen demand prediction model for the effluent from the anoxic pool; the first auxiliary variable at the time to be predicted is input into the biochemical oxygen demand prediction model for the effluent from the anoxic pool to obtain the biochemical oxygen demand of the effluent from the anoxic pool at the time to be predicted as the prediction data; According to the historical data of the nitrate-nitrogen concentration in the effluent of the anoxic pool and the second auxiliary variable, a functional relationship between the nitrate-nitrogen concentration in the effluent of the anoxic pool and the second auxiliary variable is constructed to form a prediction model for the nitrate-nitrogen concentration in the effluent of the anoxic pool; the second auxiliary variable at the time to be predicted is input into the prediction model for the nitrate-nitrogen concentration in the effluent of the anoxic pool to obtain the prediction data for the nitrate-nitrogen concentration in the effluent of the anoxic pool at the time to be predicted; Step 3: Regulate the reflow ratio; if the nitrate nitrogen concentration of the effluent from the anoxic pool is less than 1 mg / L and the biochemical oxygen demand of the effluent from the anoxic pool is less than 50 mg / L, the reflow ratio is maintained unchanged; If the nitrate nitrogen concentration of the anoxic pool effluent is less than 1 mg / L and the biochemical oxygen demand of the anoxic pool effluent is ≥50 mg / L, the reflow ratio will be increased by 10%; If the nitrate nitrogen concentration in the effluent of the anoxic pool is ≥1 mg / L, the reflux ratio will be reduced by 10%; Step 4: Delayed calibration; after the prediction data is run, 25%-50% of the actual hydraulic retention time of the anoxic tank is used as the calibration interval, the value of the prediction data corresponding to 25% of the actual hydraulic retention time of the anoxic tank is C1, and the value of the prediction data corresponding to 50% of the actual hydraulic retention time of the anoxic tank is C2; When the predicted data is the nitrate-nitrogen concentration value of the effluent from the anoxic pool, if |C1-C2|≤10%C1, the nitrate-nitrogen concentration value of the effluent from the anoxic pool remains unchanged; when C2-C1>10%C1, the compensation coefficient k is set to 1.2, and the final predicted value is 1.2*the nitrate-nitrogen concentration value of the effluent from the anoxic pool; when C1-C2>10%C1, the compensation coefficient k is set to 0.8, and the final predicted value is 0.8*the nitrate-nitrogen concentration value of the effluent from the anoxic pool; When the predicted data is the biochemical oxygen demand of the effluent from the anoxic pool, if |C1-C2|≤10%C1, the biochemical oxygen demand of the effluent from the anoxic pool remains unchanged; when C2-C1>10%C1, the compensation coefficient k is set to 0.8, and the final predicted value is 0.8*biochemical oxygen demand of the effluent from the anoxic pool; when C1-C2>10%C1, the compensation coefficient k is set to 1.2, and the final predicted value is 1.2*biochemical oxygen demand of the effluent from the anoxic pool.

2. A sewage system control method based on water quality intelligent prediction according to claim 1, characterized in that: The functional relationship between the biochemical oxygen demand of the effluent from the anoxic pool and the first auxiliary variable was constructed by a multi-form regression algorithm, and the functional relationship between the nitrate nitrogen concentration of the effluent from the anoxic pool and the second auxiliary variable was constructed by a multi-form regression algorithm.

Images