CN-121974414-A - Intelligent dosing method based on real-time adjustment of sewage flow and turbidity

CN121974414ACN 121974414 ACN121974414 ACN 121974414ACN-121974414-A

Abstract

The invention discloses an intelligent dosing method based on real-time adjustment of sewage flow and turbidity, which relates to the technical field of sewage treatment and comprises the following steps of training a prediction model through historical data and establishing a mutation feature library; collecting and preprocessing various water quality and flow parameters in real time, judging whether the water quality has severe mutation or not based on a multi-parameter index, switching a control mode according to working conditions and calculating the real-time dosing amount, executing dosing instructions and monitoring the water quality of the effluent to form feedback, updating a model and parameters on line by utilizing operation data, and circularly executing the steps to realize closed-loop intelligent control. The invention can identify the abnormal mutation state of the water quality of the inlet water in real time by introducing the intelligent decision method of the multi-parameter mutation combined criterion and the mode switching, and adaptively switch to the feedforward leading control mode with quick response according to the abnormal mutation state, thereby solving the problem that the existing method cannot timely and effectively cope with sudden and severe fluctuation of the water quality of the inlet water due to inherent hysteresis of a process link.

Inventors

LU XIAOCHEN
CAO YANFEI
WU MIN
GONG YU
XIA YANBING
GU YUNFENG
MENG XIAOJIN

Assignees

太仓武港码头有限公司

Dates

Publication Date: 20260505
Application Date: 20260106

Claims (7)

1. An intelligent dosing method based on real-time adjustment of sewage flow and turbidity is characterized by comprising the following steps: Collecting a water inflow signal, a water inflow turbidity signal and a water outflow turbidity signal in a sewage treatment process in real time; Based on the water inlet parameters acquired in real time, identifying the water quality running state, and judging whether the system is in a stable working condition or a water quality mutation event; According to the identified water quality running state, adaptively switching to a corresponding control mode, and calculating the real-time dosing amount based on an algorithm corresponding to the control mode, wherein different control modes adopt different core algorithms to calculate the dosing amount; Transmitting the calculated real-time dosing instruction to an executing mechanism to control a dosing device to dose; And on-line self-learning and updating are carried out on the control model parameters based on the system operation effect data.
2. The intelligent dosing method based on real-time adjustment of sewage flow and turbidity according to claim 1, wherein the real-time collected parameters further comprise a water inlet pH value, a water inlet temperature value and a change rate signal thereof.
3. The intelligent dosing method based on real-time adjustment of sewage flow and turbidity according to claim 1 or 2, wherein the step of identifying the water quality running state specifically comprises the following steps: Based on the real-time inflow water flow, inflow water turbidity and change rate thereof, calculating a combined mutation index which comprehensively reflects the degree of deviation of multiple parameters from a recent normal baseline; Comparing the joint mutation index to a dynamically updated recognition threshold; and when the combined mutation index continuously exceeds the identification threshold value for a preset confirmation time, judging that the water quality mutation event is entered, and otherwise, judging that the water quality mutation event is in a stable working condition.
4. The intelligent dosing method based on real-time adjustment of sewage flow and turbidity according to claim 1, wherein the step of adaptively switching control modes and calculating the real-time dosing amount comprises: When the stable working condition is identified, a first control mode is adopted, the mode uses the basic dosing quantity output by a feedforward control algorithm as a reference, and is supplemented with a feedback control algorithm based on the turbidity deviation of the effluent for fine adjustment and correction, and the two modes are overlapped to obtain the final dosing quantity; When the water quality abrupt change event is identified, the control method is immediately switched to a second control mode, wherein the mode is to superimpose an emergency compensation dosing amount predicted based on the water inlet load change trend on the basis of the basic dosing amount output by the feedforward control algorithm, and temporarily inhibit or freeze the correction effect of the feedback control algorithm in the mode.
5. The intelligent dosing method based on real-time adjustment of sewage flow and turbidity according to claim 4, wherein in the second control mode, the emergency compensation dosing amount based on the prediction of the inflow load change trend is obtained by analyzing the current inflow turbidity, inflow flow and change rate thereof, estimating the predicted pollution load in a future process reaction lag period, and combining a self-learning compensation coefficient according to the difference between the predicted load and the pre-mutation reference load.
6. The intelligent dosing method based on real-time adjustment of wastewater flow and turbidity of claim 4, further comprising a control mode exit mechanism, wherein when the system is restored to a stable working condition from a water quality abrupt change event, the second control mode maintains operation of at least one process reaction lag period, and after the effluent turbidity signal reflects the real working condition again, the first control mode is smoothly switched back.
7. The intelligent dosing method based on real-time adjustment of sewage flow and turbidity according to claim 1, wherein the step of online self-learning and updating comprises: during the running period of the stable working condition, incremental learning optimization is carried out on the feedforward control algorithm model by utilizing accumulated running data; After the water quality mutation event is processed each time, evaluating and adjusting related parameters in the emergency compensation dosing amount calculation according to the actual processing effect of the event; And according to the identification characteristics of the historical mutation events, the calculation weight of the joint mutation index and the dynamic identification threshold are adaptively optimized.

Description

Intelligent dosing method based on real-time adjustment of sewage flow and turbidity Technical Field The invention belongs to the technical field of sewage treatment, and particularly relates to an intelligent dosing method based on real-time adjustment of sewage flow and turbidity. Background In the existing sewage treatment dosing control technology, single-parameter feedback control based on effluent turbidity is a common method. The method is characterized in that the turbidity of the effluent after the sedimentation tank is monitored, and the turbidity is compared with a set target value, so that the adding amount of the medicament at the front end is reversely regulated. However, this approach has an inherent, structural disadvantage in that the detection signal has significant process hysteresis. Adding the agent into the mixed reaction, flocculating and precipitating to finally obtain the water quality change which can be detected by the effluent turbidity meter, wherein the whole process needs a quite long time. This means that the feedback information currently received by the control system reflects the process conditions and the processing effects ten to several ten minutes ago. When the quality of the incoming water is relatively stable, this hysteresis can be tolerated by conservative parameter settings. However, this drawback is greatly amplified in the face of rapid, abrupt changes in the quality of the incoming water, such as in certain industrial drainage scenarios. The feedback control system, like commanding according to the outdated information, cannot respond in time to the pollution load peaks entering the system. The regulation action is always lagged behind the actual requirement, so that the dosage is insufficient when the mutation occurs, the water quality of the effluent exceeds the standard, and the medicament is wasted due to excessive compensation after the mutation. Therefore, the following means are proposed for solving the above problems. Disclosure of Invention The invention aims to provide an intelligent dosing method based on real-time adjustment of sewage flow and turbidity, which can identify abnormal mutation states of water quality of inlet water in real time by introducing an intelligent decision method of multi-parameter mutation combined criterion and mode switching, and adaptively switch to a fast-response feedforward leading control mode according to the abnormal mutation states, so that the problem that the existing feedback control method based on the turbidity of outlet water cannot timely and effectively respond to sudden and severe fluctuation of the water quality of inlet water due to inherent hysteresis of process links is solved. In order to solve the technical problems, the invention is realized by the following technical scheme: the invention discloses an intelligent dosing method based on real-time adjustment of sewage flow and turbidity, which comprises the following steps: Collecting a water inflow signal, a water inflow turbidity signal and a water outflow turbidity signal in a sewage treatment process in real time; Based on the water inlet parameters acquired in real time, identifying the water quality running state, and judging whether the system is in a stable working condition or a water quality mutation event; According to the identified water quality running state, adaptively switching to a corresponding control mode, and calculating the real-time dosing amount based on an algorithm corresponding to the control mode, wherein different control modes adopt different core algorithms to calculate the dosing amount; Transmitting the calculated real-time dosing instruction to an executing mechanism to control a dosing device to dose; And on-line self-learning and updating are carried out on the control model parameters based on the system operation effect data. Further, the parameters collected in real time further comprise a water inlet pH value, a water inlet temperature value and a change rate signal thereof. Further, the step of identifying the water quality running state specifically comprises the following steps: Based on the real-time inflow water flow, inflow water turbidity and change rate thereof, calculating a combined mutation index which comprehensively reflects the degree of deviation of multiple parameters from a recent normal baseline; Comparing the joint mutation index to a dynamically updated recognition threshold; and when the combined mutation index continuously exceeds the identification threshold value for a preset confirmation time, judging that the water quality mutation event is entered, and otherwise, judging that the water quality mutation event is in a stable working condition. Further, the step of adaptively switching the control mode and calculating the real-time dosing amount includes: When the stable working condition is identified, a first control mode is adopted, the mode uses the basic dosing quantity output by a