Modular CSTR Fermentation Simulation Equipment and Adaptive Cooperative Control Method

Abstract

The present invention discloses a modular CSTR fermentation simulation equipment and an adaptive control method, which relate to the field of biological fermentation and process control. The equipment is composed of modules such as a detachable fermentation unit, a gas monitoring unit, and a data acquisition unit, and integrates multi-parameter sensors and compensation algorithms to achieve closed-loop control. The control method adopts a two-layer cascade architecture: the lower-level controller stabilizes the pH by adjusting the feed rate through dynamic PID; the upper-level controller dynamically optimizes the pH set value based on the gas production deviation, and forms feedback control in combination with the extreme value search model. The expert system periodically analyzes the gas production deviation, detects the load limit through micro-disturbance, and dynamically adjusts the target parameters to achieve organic load optimization. The present invention integrates multi-parameter collaborative monitoring and intelligent control models, calculates key parameters such as OLR and HRT in real time, improves processing efficiency while maintaining system stability, breaks through the hysteresis defects of traditional manual control, and provides an adaptive solution for anaerobic fermentation processes.

Description

Technical Field

[0001] The present invention relates to the field of biological fermentation and process control, and in particular to a modular CSTR fermentation simulation equipment and an adaptive coordinated control method. Background Art

[0002] Anaerobic fermentation is a process in which waste is subjected to anaerobic conditions and stably produces biogas through the metabolic activity of microorganisms. It has significant environmental benefits and economic value. In traditional anaerobic fermentation experiments, two modes are usually adopted: batch experiments or continuous experiments. Batch experiments are simple to operate and have fixed cycles, but they cannot achieve real-time dynamic monitoring of the fermentation process. While continuous experiments have the potential for long-term stable operation, their process control relies heavily on manual experience and requires manual adjustments to key process parameters such as feed rate and feed volume. This control method, which relies on manual judgment, is not only labor-intensive, but also often leads to poor system operation stability due to problems such as adjustment lag and imprecise control, which in turn affects the accuracy and repeatability of experimental results.

[0003] Existing technologies have optimized feed control and some fermentation parameters to a certain extent. For example, the automated feeding anaerobic fermentation system developed by Swedish company BIOPROCESS (publication number: WO2012005667A1) controls material inflow and outflow through pressure differentials or pumps, and employs a "first-out, last-in" order to prevent the immediate outflow of newly fed materials. This system relies on electronic balance signals to regulate valves, enabling basic control of feed and outflow rates. However, it does not address real-time monitoring and feedback control of key process parameters during the fermentation process (such as pH, alkalinity, gas composition, and gas production). Another example is a cloud-based online monitoring and control system (publication number: WO2014070099A1), which includes a fermentation unit, a data acquisition unit, and a cloud computing unit with data processing and a user interface. This system can upload online data from multiple fermentation processes to the cloud for user viewing. However, the system still has problems such as insufficient monitoring parameters and lack of online perception of core process indicators such as gas composition and flow rate, and is unable to achieve the deduction and control of key operating parameters such as organic loading rate (OLR) and hydraulic retention time (HRT).

[0004] In the field of anaerobic gas production assessment, Chinese utility model patent CN207215791U discloses an anaerobic fermentation gas assessment device that continuously and in real time monitors gas concentration and emissions to ensure the integrity and validity of experimental data. However, this device lacks integrated modules for feed control, gas information processing, and process feedback control, making it difficult to meet the high standards of real-time control required for continuous anaerobic fermentation.

[0005] In summary, although the existing technology has made certain progress in feed control, gas monitoring and data processing, there are still many problems that need to be solved. First, the functional layout of most systems is scattered, and there is a lack of effective linkage between the modules, which makes it difficult to form a synergistic effect, resulting in low overall operating efficiency of the system. Secondly, the parameters that can be monitored by existing equipment are relatively limited, and they often only focus on single indicators such as gas production or material quality, and cannot fully reflect the complex dynamic behavior and microbial reaction status in the fermentation process. At the same time, the process lacks an effective feedback control mechanism, and the operating parameters cannot be dynamically adjusted according to real-time monitoring data, making it impossible to achieve truly intelligent control. More importantly, the existing system lacks unified integration in terms of parameter integration, model calculation and process evaluation, and it is difficult to meet the current anaerobic fermentation process's comprehensive needs for precise control, data analysis and automatic optimization. Summary of the Invention

[0006] Based on this, in order to solve the problems existing in existing continuous experiments, the present invention proposes a comprehensive system that integrates multi-parameter real-time monitoring, instant feedback control and precise process control functions. The feed rate is dynamically adjusted in real time through the two core parameters of gas production and gas production change and pH or alkalinity as auxiliary parameters, thereby improving the stability of continuous experiments and the accuracy of data, reducing the labor intensity of manual operations, providing more reliable technical support for the large-scale application of biogas projects, and further promoting research and application in the field of renewable energy.

[0007] A modular CSTR fermentation simulation equipment and an adaptive coordinated control method, comprising a plurality of independently disassembled and assembled functional modules, wherein the functional modules include at least a fermentation unit, a gas composition monitoring unit, a gas flow metering unit, a data acquisition unit, and an inlet and outlet control unit;

[0008] The fermentation unit includes a bioreactor, a stirring device and a temperature control system, and each fermentation unit has a modular design, which is convenient for flexible expansion and combination;

[0009] The fermentation unit is connected to the data acquisition unit and is used to transmit at least one online real-time monitoring data of the fermentation process to the feeding and discharging control unit to achieve remote visual viewing and automatic control;

[0010] The gas component monitoring unit is arranged between the fermentation unit and the gas flow metering unit, and is used to perform component analysis on the fermentation gas and transmit the analysis data to the data acquisition unit, while delivering the gas to the gas flow metering unit;

[0011] The gas flow metering unit is used to measure the gas production flow and send the measurement data to the data acquisition unit to achieve full process data monitoring and closed-loop control support.

[0012] The modular CSTR fermentation simulation equipment is a multi-channel configuration, and therefore usually requires more than one fermentation or gas component and gas flow metering unit.

[0013] Furthermore, the bioreactor is equipped with a feeding unit and a discharging unit for realizing the quantitative addition of reactants and the automatic discharge of fermentation products; the equipment supports the parallel operation of no less than 6 fermentation units and can be used for simultaneous comparative tests of multiple reactors.

[0014] Furthermore, the stirring device in the fermentation unit includes a stirring shaft driven by a motor, which can mix the materials in the reactor.

[0015] Furthermore, the gas flow measurement device is used to achieve automatic real-time gas standardization, measuring factors including temperature, pressure and water vapor content, and accurately monitoring small-scale gas flow.

[0016] The data acquisition unit includes but is not limited to sensors for collecting parameters such as pH, temperature, pressure, gas composition, oxidation-reduction potential (ORP), alkalinity, etc., and can realize online real-time data acquisition of one or more fermentation processes in at least one fermentation unit and / or data acquisition unit. The data acquisition unit is equipped with hardware and software interfaces compatible with multiple open platforms to support flexible integration and remote expansion of the system.

[0017] Furthermore, the data acquisition unit has built-in automatic real-time temperature and pressure normalization compensation algorithm, water vapor removal compensation algorithm and flushing gas overestimation removal compensation algorithm, and configures the compensation algorithm according to actual needs to output experimental data that meets different data.

[0018] Furthermore, the data acquisition unit has capabilities such as database, file storage and user interface.

[0019] Furthermore, the data acquisition unit is capable of transmitting data online in real time.

[0020] Furthermore, the online real-time data includes but is not limited to organic loading rate, hydraulic retention time, gas composition, and gas production rate.

[0021] Furthermore, the system configuration further comprises at least one sensor for measuring pH, temperature, pressure, gas composition, ORP, alkalinity, VFA, biodegradable organic matter or fermentation metabolites.

[0022] Furthermore, the organic load (OL) is calculated each time a load is added to the system and the average value is taken for report generation. Depending on whether the feed is solid or liquid, the organic load is calculated online as follows:

[0023]

[0024] Where OL is the organic load, which is the amount of organic matter added at one time or cumulatively; F is the feed amount, which is the total amount added at one time or cumulatively; Conc is the material concentration, which is the organic matter concentration of the added material.

[0025] Furthermore, the organic loading rate (OLR) is used to quantify the daily feed volume per unit volume of the fermenter (solid raw materials are represented by VS, liquid raw materials are represented by COD), and the unit is g VS L -1 d -1 or g COD L -1 d -1 OLR is calculated from the fermenter volume, feed concentration, and total feed volume, and is therefore the best parameter for adjusting feed and evaluating reactor performance.

[0026]

[0027] Among them, OLR is the organic loading rate, OL is the organic load, which means the amount of organic matter added once or cumulatively, V is the volume of the fermentation tank, t is the feeding time, Indicates the feeding interval time, which is not disturbed by the user at different feeding times and can accurately obtain data.

[0028] Furthermore, the hydraulic retention time (HRT) that can be calculated online is calculated online by the ratio of the fermentation tank volume to the feed amount:

[0029]

[0030] Where V represents the volume of the fermenter, t is the feeding time, Indicates the feeding interval time, F is the feeding amount, which is the total amount added at one time or cumulatively;

[0031] Furthermore, in this system, organic loading rate (OLR) and hydraulic retention time (HRT) can be calculated in real time and displayed alongside normalized gas flow rates. To reduce data processing and transmission, users can select different time intervals for detailed data analysis or longer time intervals for an overall process overview. Other key process parameters, such as specific gas production (SGP) and gas yields for both total and organic feeds, can also be calculated and stored.

[0032] Furthermore, selecting an appropriate hydraulic retention time (HRT) can achieve high methane yields. In a CSTR, a short HRT can flush the microbial population out of the reactor, leading to fermentation failure. The recommended HRT for a CSTR is 20 days to reduce the risk of microbial washout. A longer HRT increases the time microorganisms have to degrade the material, which can increase biogas production, but also reduces biogas production efficiency. Therefore, balancing gas production and HRT is crucial.

[0033] The preferred time for HRT is 20-30 days when OLR>3, and 15-20 days when OLR≦3.

[0034] Furthermore, the system analyzes the gas components produced by the bioreactor through a gas composition monitoring unit, measures the gas production volume through a gas flow metering unit, and collects the above information through a data acquisition unit to calculate the organic loading rate (OLR) and hydraulic retention time (HRT). The system then optimizes the parameters through piecewise linear interpolation based on historical data.

[0035] The data acquisition unit and the inlet and outlet control unit realize the joint dynamic adjustment of multiple parameters such as gas production, methane concentration, pH or alkalinity;

[0036] The control model for the operation of the data acquisition unit and the feed and discharge control unit includes an upper-level controller, a lower-level controller, an anaerobic process dynamic module and an extreme value search model; among them, the feed control unit adopts a two-level control architecture, in which the upper-level controller and the lower-level controller work together to achieve fine control of the reaction process.

[0037] The feed control unit monitors changes in key fermentation parameters (temperature, pH, stirring speed, and feed rate) in real time, allowing the system to determine whether the fermentation is normal. If an anomaly is detected, the system automatically responds by adjusting stirring speed, feed rate, and other parameters based on the calculated organic loading rate, evenly distributing the required organic matter content over a longer feeding cycle to ensure a stable fermentation environment.

[0038] Furthermore, the system receives a difference e2 between a set gas production rate value and an actual gas production rate value through an upper-level controller, and the controller outputs a pH set value based on the difference; a difference operation is performed between the pH set value and the actual pH measurement value collected from the fermentation unit to obtain a pH deviation value e1, and the pH deviation value is input to a lower-level controller; the lower-level controller outputs a feed rate based on the pH deviation value e1 to adjust the anaerobic process; the fermentation unit operates based on the feed rate, and outputs parameters such as the actual gas production rate and the gas production rate change rate; at the same time, the actual gas production rate and the gas production rate change rate and other parameters are fed back to the expert decision model, and the model dynamically adjusts the gas production rate set value based on the feedback parameters and the optimization strategy; a closed-loop control system is formed to achieve dynamic optimization and regulation of the anaerobic fermentation process.

[0039] Furthermore, the upper-level fuzzy controller receives the measured gas production rate from the reactor and compares it with the set target gas production rate, generating an error e2. Based on fuzzy rules, it outputs a new pH setpoint (or a correction to the original setpoint) and transmits it to the lower-level controller. The lower-level PID controller compares this pH setpoint with the actual pH measurement in the reactor, generating an error e1, and outputs a control signal to adjust the feed pump speed (feed flow rate). The reactor operates at the new feed rate, and its gas production rate and pH status are fed back via sensors. The pH sensor has a fast response, reflecting real-time changes in the fermentation broth's acidity and alkalinity. The gas flow rate sensor accumulates gas production volume to calculate the gas production rate, providing slower feedback. The upper-level controller also has manually defined safety thresholds (pH lower limit 6.8, upper and lower limits for gas production rate, etc.) to trigger protection logic in the event of an abnormality. When an abnormality is detected (such as pH falling below the threshold or excessive gas production deviation), the fuzzy control module activates the appropriate control rules, such as forcing a reduction in feed flow. The entire framework forms a closed-loop system, in which the outer loop (gas production control) and the inner loop (pH control) are nested and work together to achieve multi-objective control of the CSTR fermentation process (maintaining a stable pH and optimizing gas production).

[0040] The adaptive collaborative control method of the modular CSTR fermentation simulation equipment is configured such that the feed control unit can also adjust the feed rate based on pH and gas production rate. The control system is divided into a two-layer cascade control system, including a lower controller and an upper controller:

[0041] (1) The lower controller is used to maintain the pH value within a set range;

[0042] Control output: adjust feed speed;

[0043] Proportional control: u1(t) = u 10 + K1× e1(t);

[0044] Among them, u1(t): feed speed adjustment value, u10 : initial bias value of feed rate, K1: lower stage proportional gain, e1(t): error between actual pH value and set value (e1(t) = pH - pH a );

[0045] (2) The upper controller is used to optimize the biogas production of the system;

[0046] Control output: Dynamically adjust the pH set value of the lower controller;

[0047] Proportional-integral control: u2(t) = u 20 + K2× e2(t) + K 2i ×∫e2(t)dt;

[0048] Where, u2(t): the set value adjustment of the lower pH, u 20 : Initial pH setting value, K2: Upper proportional gain, K 2i : Upper integral gain, e2(t): Error between gas production target value and actual value (e2(t)=QQ a , where Q is the target gas production, Q a is the actual gas production).

[0049] The lower-level controller uses a variable proportional gain PID control algorithm to adjust the feed flow in real time to maintain stable pH within the set range. The proportional gain KP of the PID controller is dynamically adjusted according to the change of ΔpH.

[0050] The upper controller, based on the gas production rate Q a The deviation e2 from the target gas production rate Q is dynamically adjusted to adjust the target pH and feed control strategy;

[0051] Furthermore, the adaptive collaborative control method of the modular CSTR fermentation simulation equipment comprises:

[0052] An expert decision-making feedback module periodically corrects the gas production rate Q based on system operating trends, load changes, and disturbance responses. This correction actively detects the reactor load limit by applying small positive disturbances without significantly disturbing the system stability, thereby achieving improved processing capacity and extreme value optimization control.

[0053] The expert decision feedback module combines the rule-driven extreme value search model to achieve dynamic optimization, including:

[0054] 1. Extreme value search strategy:

[0055] When the pH set value reaches the lower limit (such as 6.95), the system automatically reduces the feed rate to prevent the pH from continuing to drop. At the same time, it dynamically balances the gas production efficiency and system stability by adjusting the upper and lower limits (e2_max and e2_min) of the gas production rate.

[0056] During the startup phase (OLR < 1.5 g VS L -1 d -1 ), set a higher e2_max (such as 0.10 L gas · L reactor -1 ·d -1 ) to accelerate the enrichment of microbial communities and increase the initial gas production rate;

[0057] When approaching the maximum load of the system (OLR>3.0g VS L -1 d -1 ), reduce e2_max to 0.04 L gas ·L reactor -1 ·d -1 And narrow the e2_min range (such as ±0.05) to reduce the feeding "pushing authority" and avoid VFA accumulation.

[0058] 2. Dynamic threshold adjustment mechanism:

[0059] The dynamic threshold adjustment mechanism introduced in this model is an adaptive control strategy based on real-time system status. It dynamically updates the upper and lower limits (e2_max and e2_min) of the gas production rate difference to balance gas production efficiency and system stability. Based on the coordinated response of real-time gas production rate and pH, its core logic is as follows:

[0060]

[0061] Where, β=0.002(mg / L) -1 , indicating that for every 1 mg / L increase in VFA concentration, e2_min increases by 0.002, pushing the system to increase feed to alleviate VFA accumulation;

[0062] γ = 0.15, which means that for every 1 unit that the pH deviates from the set value, e2_min increases by 0.15, and the acid-base imbalance is corrected by adjusting the feed rate.

[0063] 3. Transient balance control strategy:

[0064] The transient equilibrium control strategy introduced in the model is a dynamic control method for the startup, load mutation, and external disturbance recovery stages of the anaerobic fermentation system, aiming to quickly stabilize key parameters (such as pH, VFA concentration, and gas production rate) to prevent instability or efficiency loss.

[0065] When the system is in a high load transient state (such as OLR>2.5g VS L -1 d -1 ), prioritize maintaining the lower pH set point (6.8-7.0), and balance gas production and stability through the following strategies:

[0066]

[0067] Where λ = 0.3 is the attenuation coefficient,

[0068] By real-time monitoring of pH, VFA, and gas production rate, when VFA>2000 mg / L, the feed rate was forced to be reduced to 50% of the baseline value.

[0069] The control method includes the following strategies:

[0070] The system pH value is collected every 8 hours and the pH deviation e1 is calculated. The K in the PID controller is automatically updated according to the e1 amplitude. p value;

[0071] The gas production rate Q is collected every 24 hours a The deviation e2 from the target gas production rate Q is calculated to determine the current processing load and reaction efficiency;

[0072] Based on the interval where the deviation e2 is located, the expert module outputs correction suggestions and dynamically adjusts the target gas production rate Q and target pH to achieve system adaptive load regulation and maximize gas production performance.

[0073] Furthermore, the adaptive collaborative control method of the modular CSTR fermentation simulation equipment, the lower-level proportional gain K1 in the lower-level controller is adjusted according to the current gas production rate Gas production rate at the previous moment The feed rate F(t) can be adjusted by adaptively changing the difference D(t) between them.

[0074]

[0075] The specific strategy of the expert database decision model is as follows:

[0076] (1) When hour,

[0077] The output value , at this time you can increase the feed;

[0078] (2) When hour,

[0079] F(t) K 12 × , at this time, it is preferred to feed according to the original feed amount;

[0080] (3) When When the last monitoring cycle D hour,

[0081] F(t) K 13 × At this time, the feeding should be stopped and the changes in gas production should be observed for 1 day.

[0082] in, This is the experience value after system debugging.

[0083] The adaptive collaborative control method of the modular CSTR fermentation simulation equipment of this embodiment is configured such that the feed control unit can also adjust the feed rate based on alkalinity and gas production rate and is divided into a two-layer cascade control system, wherein:

[0084] Lower level controller:

[0085] Function: Maintain the total alkalinity (TA) in the reactor within the dynamic setting range to prevent system instability caused by acid accumulation.

[0086] Control output: adjust the dosing rate of alkaline additives (such as sodium bicarbonate) or the dilution reflux ratio.

[0087] Adaptive proportional-integral control:

[0088]

[0089] Where, u1(t): Alkaline additive injection rate adjustment value, e1(t)= TA actual -TA set : The deviation between the real-time alkalinity value and the set value, K1 and K 1i : Dynamically adjusted proportional and integral gains, adaptively updated according to the rate of change of alkalinity (ΔTA / Δt).

[0090] Upper controller:

[0091] Function: By dynamically optimizing the alkalinity setting value (TA set ), maximize the biogas yield (Q gas )

[0092] Control output: Real-time adjustment of the alkalinity setpoint of the lower controller

[0093] Control algorithm (model predictive optimization):

[0094]

[0095] Among them, TA opt: The optimal empirical value of alkalinity based on historical data, λ: Stability weight factor to prevent drastic fluctuations in the set value.

[0096] The model searches within the range of TA to maximize the objective function and obtain the optimal TA value.

[0097] The expert decision feedback module (alkalinity-load collaborative optimization) includes:

[0098] 1. Dynamic load increase strategy:

[0099] A step-type organic load disturbance (increase ≤ 5%) was applied every 12 or 24 hours, and the alkalinity buffering capacity (β = ΔVFA / ΔTA) and gas production rate response sensitivity (S = ΔQ gas / ΔOLR).

[0100] According to the judgment whether the "if , then return to the previous stable load" condition, then the load is allowed to continue to increase.

[0101] 2. Extreme value locking mechanism:

[0102] When it is detected that the rate of change of gas production rate approaches zero, that is, D(t)≈0, the start strategy is to fine-tune TA in the range of ±200 mg / L with a step size of 50 mg / L. set , and through the gas production activation energy model (E α = f(TA, VFA)) to determine the optimal alkalinity operating point.

[0103] Beneficial effects of the present invention:

[0104] A key innovation of this invention lies in the integration of a dynamic regulation system based on a control model that automatically calculates and visualizes key process parameters in real time, enabling real-time dynamic adjustment of feed rates. The system is compatible with both manual and automatic feeding modes, provides real-time monitoring of key indicators such as pH, gas methane composition, and gas production, and automatically calculates the organic loading rate (OLR) and hydraulic retention time (HRT) of the CSTR reactor.

[0105] Based on a two-layer control loop centered on pH or alkalinity and gas production rate, the system introduces an extreme value search model to construct a feedback control mechanism. This can dynamically identify the optimal feeding strategy, extract key characteristic parameters, enhance prediction accuracy, and model daily gas production and daily gas production changes, thereby enhancing the system's adaptability under variable operating conditions. This mechanism can effectively capture the complex dynamic characteristics of anaerobic fermentation systems and support refined feed adjustments and operational optimization. Through this control model, the system can optimize the feed rate based on real-time gas production, dynamically adapt to changes in fermentation status, and provide operators with an intuitive control basis through the dynamic visualization window output by the system.

[0106] On this basis, this invention emphasizes the pursuit of "high-load yet stable operation." In practice, the goal of maximizing OLR should be to find a Pareto optimal solution between processing capacity, gas production efficiency, and system stability. Therefore, this system combines multi-parameter online monitoring methods (such as VFA real-time sensors and CH4 content analyzers) with control model algorithms (such as the ADM1 module or extreme value search model) to ensure stable and efficient operation under complex nonlinear dynamics, avoiding the risk of instability caused by blindly pursuing theoretical limit loads.

[0107] This invention achieves intelligent automated operation, significantly reducing manual intervention, lowering operating and management costs, and improving experimental efficiency. Compared to traditional control methods that rely on experience, this system, through the coupling of a two-tier control architecture and intelligent models, achieves precise regulation and process parameter optimization of the anaerobic fermentation process, significantly improving the overall system performance and operational reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0108] Figure 1 is a system block diagram of the present invention;

[0109] Figure 2 This is the organic load rate control strategy of Example 2 of the present invention;

[0110] Figure 3 Graph showing daily gas production and daily gas production change in Example 2 of the present invention;

[0111] Figure 4 This is a schematic diagram of the control model of Example 2 of the present invention;

[0112] Figure 5 The figure shows the daily gas production and gas production fluctuation of the control group 1 of the present invention;

[0113] Figure 6 Graph showing daily gas production and gas production fluctuations for control group 2 of the present invention. DETAILED DESCRIPTION

[0114] To facilitate understanding of the present invention, the present invention will be described more fully below with reference to the accompanying drawings. The accompanying drawings illustrate preferred embodiments of the present invention. However, the present invention may be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided to provide a more thorough and comprehensive understanding of the present disclosure.

[0115] like Figure 1As shown, a modular CSTR fermentation simulation equipment and an adaptive coordinated control method include multiple functional modules that can be independently disassembled and combined. The functional modules include at least a fermentation unit, a gas component monitoring unit, a gas flow metering unit, a data acquisition unit, and an inlet and outlet control unit.

[0116] The fermentation unit includes a bioreactor, a stirring device and a temperature control system, and each fermentation unit has a modular design, which is convenient for flexible expansion and combination.

[0117] Figure 4 The control system described in is a two-level control with a supervisory loop in the form of a rule-based system, consisting of a lower-level controller and a higher-level controller:

[0118] (1) The lower controller is used to maintain the pH value within a set range;

[0119] Control output: adjust feed speed;

[0120] Proportional control: u1(t) = u 10 + K1× e1(t);

[0121] Among them, u1(t): feed speed adjustment value, u 10 : initial bias value of feed rate, K1: lower stage proportional gain, e1(t): error between actual pH value and set value (e1(t) = pH - pH a );

[0122] (2) The upper controller is used to optimize the biogas production of the system;

[0123] Control output: Dynamically adjust the pH set value of the lower controller;

[0124] Proportional-integral control: u2(t) = u 20 + K2× e2(t) + K 2i ×∫e2(t)dt;

[0125] Among them, u2(t): the set value adjustment of the lower pH, u2: the initial pH set value, K2: the upper proportional gain, K 2i : Upper integral gain, e2(t): Error between gas production target value and actual value (e2(t) = Q - Q a , where Q is the target gas production, Q a is the actual gas production).

[0126] The lower-level controller uses a variable proportional gain PID control algorithm to adjust the feed flow in real time to maintain stable pH within the set range. The proportional gain KP of the PID controller is dynamically adjusted according to the change of ΔpH.

[0127] The upper controller, based on the gas production rate Q a The deviation e2 from the set target gas production rate Q is dynamically adjusted to adjust the target pH and feed control strategy;

[0128] The adaptive collaborative control method of the modular CSTR fermentation simulation equipment comprises:

[0129] An expert decision-making feedback module periodically corrects Q based on system operating trends, load changes, and disturbance responses. This correction actively detects the reactor load limit by applying small positive disturbances without significantly disturbing the system stability, thereby achieving improved processing capacity and extreme value optimization control.

[0130] The system pH value is collected every 8 hours and ΔpH is calculated. The KP value in the PID controller is automatically updated according to the ΔpH amplitude.

[0131] The gas production rate Q is collected every 24 hours a The deviation e2 from the target gas production rate Q, Q and pH are calculated to achieve system adaptive load regulation and maximize gas production performance.

[0132] Each outer loop control cycle (taking daily as an example):

[0133] (a) The gas production rate Q(t-1) of the previous day and the actual gas production rate Q(t) of the current day are read, and the gas production rate change D(t) is calculated and adaptively changed to adjust the feed rate F(t).

[0134] D(t)=

[0135] (b) Input the difference D(t) into the expert decision model and infer the new pH adjustment value ΔpH based on the expert decision model. For example, if e2 is positive and large, the output is positive ΔpH s (Increase the set pH); if e2 is negative and the absolute value is large, the output is negative ΔpH s (Lower the set pH); when the error is small, the output is close to zero.

[0136] The specific strategy of the expert database decision model is as follows:

[0137] (1) When hour,

[0138] The output value , at this time you can increase the feed;

[0139] (2) When hour,

[0140] F(t) K 12 × , at this time, it is preferred to feed according to the original feed amount;

[0141] (3) When When the last monitoring cycle D hour,

[0142] F(t) K 13 × At this time, the feeding should be stopped and the changes in gas production should be observed for 1 day.

[0143] in, This is the experience value after system debugging.

[0144] (c) Update the pH of the lower-level controller: pH(t) = pH(t-1) + ΔpH (and limit the pH setting to no less than the safety lower limit of 6.8).

[0145] (d) Run the lower-level control in each inner-loop control cycle (e.g., every half day): read the current actual pH value and compare it with the actual pH value to obtain the deviation e1.

[0146] (e) Calculate the lower-level control output, i.e., the feed rate adjustment u1(t). The lower-level PID controller calculates the control action based on the deviation e1. Here, proportional control is usually used: u1(t) = u 10 + K1× e1(t).

[0147] (f) u1(t) is sent to the actuator to adjust the feed rate. If u1(t) is positive and large, it indicates a need to increase the feed rate; if it is negative, it indicates a need to decrease the feed rate. Feed rate adjustment can be achieved by adjusting the pump speed or opening a bypass.

[0148] (g) The reactor operates under the new feed conditions for a period of time, with sensors continuously monitoring parameters such as pH and gas production. Before entering the next control cycle, the system determines whether there have been any abnormal disturbances: checking whether the pH has ever fallen below 6.8 during this cycle, whether the gas production rate has fluctuated dramatically beyond a set threshold, or whether other monitored parameters (such as alkalinity) have exceeded their limits. If an abnormality is detected, the system enters the disturbance handling branch: executing predetermined control actions such as temporarily shutting down the feed or extending the control interval for this cycle, and then skipping to (i). If there are no abnormalities, the system continues directly.

[0149] (h) Disturbance handling: Execute the corresponding control strategy (feed reduction, feed stop or other operations) according to the type of disturbance, and continue monitoring until the indicators return to normal or reach the predetermined time, and then resume normal control.

[0150] (i) Update historical data: Record the gas production, pH, feed rate, etc. for future analysis and parameter adjustment. Then, wait for the next control cycle to begin, return to step (a), and repeat the cycle.

[0151] The above process implements a complete closed-loop control cycle. During continuous operation, the upper-level fuzzy control adjusts the gas production target each cycle, gradually bringing the system closer to the optimal load. The lower-level control stabilizes the pH in real time to ensure that the system does not exceed its limits. Meanwhile, the monitoring module is always on guard, and if it detects any abnormal changes in feed or load, it will immediately take measures to bring the system back to a safe range. The above process demonstrates how this optimized control strategy switches between normal regulation and abnormal correction, thereby ensuring the continuous and efficient operation of the CSTR reactor.

[0152] [Example 1]

[0153] In a specific embodiment, a continuous feeding experiment is conducted, and the material in the feed tank is pretreated. The experimental device includes multiple functional modules that can be independently disassembled and assembled, and the functional modules include at least a fermentation unit, a gas composition monitoring unit, a gas flow metering unit, a data acquisition unit, and an inlet and outlet control unit. The fermentation unit of this embodiment includes a bioreactor, a temperature control system, and a stirring device. The bioreactor is a CSTR reactor, and the data acquisition unit is capable of collecting parameters such as pH, temperature, pressure, gas composition, ORP, alkalinity, and VFA.

[0154] The adaptive collaborative control method of the modular CSTR fermentation simulation equipment of this embodiment is configured such that the feed control unit can also adjust the feed rate based on alkalinity and gas production rate and is divided into a two-layer cascade control system, wherein:

[0155] Lower level controller:

[0156] Function: Maintain the total alkalinity (TA) in the reactor within the dynamic setting range to prevent system instability caused by acid accumulation.

[0157] Control output: adjust the dosing rate of alkaline additives (such as sodium bicarbonate) or the dilution reflux ratio.

[0158] Adaptive proportional-integral control:

[0159]

[0160] Where, u1(t): Alkaline additive injection rate adjustment value, e1(t)= TAactual -TA set : The deviation between the real-time alkalinity value and the set value, K1 and K 1i : Dynamically adjusted proportional and integral gains, adaptively updated according to the rate of change of alkalinity (ΔTA / Δt).

[0161] Upper controller:

[0162] Function: By dynamically optimizing the alkalinity setting value (TA set ), maximize the biogas yield (Q gas )

[0163] Control output: Real-time adjustment of the alkalinity setpoint of the lower controller

[0164] Control algorithm (model predictive optimization):

[0165]

[0166] Among them, TA opt : The optimal empirical value of alkalinity based on historical data, λ: Stability weight factor to prevent drastic fluctuations in the set value.

[0167] The model searches within the range of TA to maximize the objective function and obtain the optimal TA value.

[0168] The expert decision feedback module (alkalinity-load collaborative optimization) includes:

[0169] 1. Dynamic load increase strategy:

[0170] A step-type organic load disturbance (increase ≤ 5%) was applied every 12 or 24 hours, and the alkalinity buffering capacity (β = ΔVFA / ΔTA) and gas production rate response sensitivity (S = ΔQ gas / ΔOLR).

[0171] According to the judgment whether the "if , then return to the previous stable load" condition, then the load is allowed to continue to increase.

[0172] 2. Extreme value locking mechanism:

[0173] When it is detected that the rate of change of gas production rate approaches zero, that is, D(t)≈0, the start strategy is to fine-tune TA in the range of ±200 mg / L with a step size of 50 mg / L. set , and through the gas production activation energy model (E α = f(TA, VFA)) to determine the optimal alkalinity operating point.

[0174] [Example 2]

[0175] In a specific embodiment, a continuous feeding experiment is conducted, and the material in the feed tank is pretreated. The experimental device includes multiple functional modules that can be independently disassembled and assembled, and the functional modules include at least a fermentation unit, a gas composition monitoring unit, a gas flow metering unit, a data acquisition unit, and an inlet and outlet control unit. The fermentation unit of this embodiment includes a bioreactor, a temperature control system, and a stirring device. The bioreactor is a CSTR reactor, and the data acquisition unit is capable of collecting parameters such as pH, temperature, pressure, gas composition, ORP, alkalinity, and VFA.

[0176] In this example, this method has been shown to significantly improve the hydrolysis efficiency of the material. During the fermentation process, the material that has undergone hydrothermal pretreatment has a faster gas production rate and a higher feed response rate. In this example, the pretreated material is Miscanthus sinensis stems (air-dried for 30 days and then crushed to 20 mesh) and leaves (cut into 1-1.5 cm pieces) and stored at 4°C for later use.

[0177] The modular CSTR fermentation simulation equipment is a multi-channel configuration, and therefore usually requires more than one fermentation or gas component and gas flow metering unit.

[0178] The system consists of a fermentation unit (CSTR reactor), a gas composition monitoring unit, a gas flow meter unit, and a data acquisition and control unit. The core components of the fermentation unit are the bioreactor, condenser, and stirring paddle. The data acquisition and control unit integrates sensors for measuring pH, temperature, pressure, gas composition, ORP (oxidation-reduction potential), alkalinity, VFA (volatile fatty acids), biodegradable organic matter, and fermentation metabolites.

[0179] Organic loading rate in the startup phase: 0.5g VS L -1 d -1 The material is fed into the CSTR reactor from the feed device at an organic loading rate of 100%. The stirring device is started to mix the material. After the material is fully in contact with the microorganisms in the inoculated mud, gas production begins.

[0180] Improvement of organic loading rate: As the fermentation progresses, the organic loading rate is gradually increased, and the microbial community suitable for the fermentation material is enriched to achieve a stable organic loading rate.

[0181] Solid residence time: set to 20 days.

[0182] During the reaction, unknown quantities are calculated: Organic Loading Rate (OLR): The amount of organic matter input, known from the set feed rate and feedstock characteristics. Normalized Gas Volume: The system calculates the gas volume under standard conditions using a normalization algorithm for temperature, pressure, and humidity. Concentration of Key Metabolites: The production rate of metabolites is inferred based on changes in gas composition and metabolic models.

[0183] This interactive calculation of known and unknown quantities enables the system to achieve precise control and optimization, automatically adjusting parameters to ensure the stability and efficiency of the fermentation process.

[0184] OL

[0185] Where OL is the organic load, which is the amount of organic matter added at one time or cumulatively; F is the feed amount, which is the total amount added at one time or cumulatively; Conc is the material concentration, which is the organic matter concentration of the added material.

[0186]

[0187] Among them, OLR is the organic loading rate, OL is the organic load, which means the amount of organic matter added once or cumulatively, V is the volume of the fermentation tank, t is the feeding time, Indicates the feeding interval time.

[0188]

[0189] Where V represents the volume of the fermenter, t is the feeding time, Indicates the feeding interval time, F is the feeding amount, which is the total amount added at one time or cumulatively;

[0190] When fermentation is abnormal, the system automatically calculates and re-establishes the feeding strategy by adjusting the material residence time and the organic loading rate.

[0191] First, historical data were collected to obtain data points of system performance indicators (such as gas production rate, methane content, and volatile fatty acid accumulation) under different OLRs.

[0192] The adaptive collaborative control method of the modular CSTR fermentation simulation equipment of this embodiment is configured such that the feed control unit can also adjust the feed rate based on pH and gas production rate and is divided into a two-layer cascade control system, including: a lower controller and an upper controller:

[0193] (1) The lower controller is used to maintain the pH value within a set range;

[0194] Control output: adjust feed speed;

[0195] Proportional control: u1(t) = u 10 + K1× e1(t);

[0196] Among them, u1(t): feed speed adjustment value, u 10 : initial bias value of feed rate, K1: lower stage proportional gain, e1(t): error between actual pH value and set value (e1(t) = pH - pH a );

[0197] (2) The upper controller is used to optimize the biogas production of the system;

[0198] Control output: Dynamically adjust the pH set value of the lower controller;

[0199] Proportional-integral control: u2(t) = u 20 + K2× e2(t) + K 2i ×∫e2(t)dt;

[0200] Where, u2(t): the set value adjustment of the lower pH, u 20 : Initial pH setting value, K2: Upper proportional gain, K 2i : Upper integral gain, e2(t): Error between gas production target value and actual value (e2(t) = Q - Q a , where Q is the target gas production, Q a is the actual gas production).

[0201] The lower-level controller uses a variable proportional gain PID control algorithm to adjust the feed flow in real time to maintain stable pH within the set range. The proportional gain KP of the PID controller is dynamically adjusted according to the change of ΔpH.

[0202] The upper controller dynamically adjusts the target pH and feed control strategy based on the deviation e2 between the gas production rate Q and the target gas production rate Q;

[0203] Furthermore, the adaptive collaborative control method of the modular CSTR fermentation simulation equipment comprises:

[0204] An expert decision-making feedback module periodically corrects Q based on system operating trends, load changes, and disturbance responses. This correction actively detects the reactor load limit by applying small positive disturbances without significantly disturbing the system stability, thereby achieving improved processing capacity and extreme value optimization control.

[0205] The control method comprises the following steps:

[0206] The control mechanism adopts a dual-loop structure:

[0207] Lower controller (proportional control): adjusts the feed rate in real time based on pH error,

[0208] The formula is u1(t)=u 10 +K1xe1(t);

[0209] Upper controller (proportional integral control): adjusts the pH set value according to the biogas production error,

[0210] The formula is

[0211] Among them, K1 is adaptively adjusted according to the gas production rate difference D(t), and the threshold range is D_max=0.3, D_min=-0.45.

[0212] The expert decision feedback module combines the rule-driven extreme value search model to achieve dynamic optimization, including:

[0213] 1. Extreme value search strategy:

[0214] When the pH set value reaches the lower limit (such as 6.95), the system automatically reduces the feed rate to prevent the pH from continuing to drop, and dynamically balances the gas production efficiency and system stability by adjusting the e2_max and e2_min thresholds;

[0215] During the startup phase (OLR < 1.5 g VS L -1 d -1 ), set a higher e2_max (such as 0.10 L gas · L reactor -1 ·d -1 ) to accelerate the enrichment of microbial communities and increase the initial gas production rate;

[0216] When approaching the maximum load of the system (OLR>3.0g VS L -1 d -1 ), reduce e2_max to 0.04 L gas ·L reactor -1 ·d -1 And narrow the e2_min range (such as ±0.05) to reduce the feeding "pushing authority" and avoid VFA accumulation.

[0217] 2. Dynamic threshold adjustment mechanism:

[0218] The dynamic threshold adjustment mechanism introduced in this model is an adaptive control strategy based on real-time system status. It dynamically updates the upper and lower limits (e2_max and e2_min) of the gas production rate difference to balance gas production efficiency and system stability. Based on the coordinated response of real-time gas production rate and pH, its core logic is as follows:

[0219]

[0220] Where, β=0.002(mg / L) -1 , indicating that for every 1 mg / L increase in VFA concentration, e2_min increases by 0.002, pushing the system to increase feed to alleviate VFA accumulation;

[0221] γ = 0.15, which means that for every 1 unit that the pH deviates from the set value, e2_min increases by 0.15, and the acid-base imbalance is corrected by adjusting the feed rate.

[0222] 3. Transient balance control strategy:

[0223] The transient equilibrium control strategy introduced in the model is a dynamic control method for the startup, load mutation, and external disturbance recovery stages of the anaerobic fermentation system, aiming to quickly stabilize key parameters (such as pH, VFA concentration, and gas production rate) to prevent instability or efficiency loss.

[0224] When the system is in a high load transient state (such as OLR>2.5g VS L -1 d -1 ), prioritize maintaining the lower pH set point (6.8-7.0), and balance gas production and stability through the following strategies:

[0225]

[0226] Where λ = 0.3 is the attenuation coefficient,

[0227] By real-time monitoring of pH, VFA, and gas production rate, when VFA>2000 mg / L, the feed rate was forced to be reduced to 50% of the baseline value.

[0228] The specific control process is as follows:

[0229] If the VFA concentration suddenly rises to 2600 mg / L due to excessive feed and the pH drops to 6.75, the rule-driven model is triggered, the feed rate is reduced by 50%, and the rapid addition of buffer (such as sodium bicarbonate) is started to increase the alkalinity.

[0230] The model sets a lower limit for the minimum gas production rate according to e2_min(t)=0.002×2600+0.15×(6.75-7.0)=5.1625; and gradually recovers the gas production rate by optimizing the pH setting value.

[0231] When VFA < 800 mg / L and pH is stable above 6.95, e2_max is gradually relaxed and normal feeding is resumed.

[0232] The expert decision feedback module also combines real-time monitoring data from the data acquisition and control unit, such as ORP, volatile fatty acid content and pH value, to dynamically adjust feed parameters.

[0233] In this example, in order to further verify the advantages of the present invention, a comparative experiment was set up. In addition to using the continuous anaerobic fermentation system and control method of the present invention, two groups of control experiments were set up:

[0234] One group used the traditional method of adjusting feed amount based only on experience (control group 1);

[0235] The other group used a method based on ordinary linear regression model to predict gas production and adjust feed (control group 2).

[0236] During the experiment, other conditions such as fermentation raw materials, reactor type, and ambient temperature were kept consistent. As for the key indicator of gas production, the system of the present invention increased with the organic loading rate from 1.0 g VS L - 1 d -1 Gradually increase to 3.5g VS L -1 d -1 The total gas production showed a steady growth trend. When the organic loading rate was 2 g VS L -1 d -1 When the total gas production reached 2 times of the initial stage; at 3.5 g VS L -1 d -1 In the high load stage, the gas production is close to the theoretical maximum. However, after the organic loading rate is increased, the gas production of the control group 1 fluctuates violently. -1 d -1 When the load is increased, the gas production is only increased by 1.2 times, and the gas production is stagnant or even decreases during the subsequent load increase process (such as Figure 5 As shown); although the control group 2 showed some growth, the growth rate was significantly lower than that of the system of the present invention. -1 d -1 When the gas production of the system of the present invention is only 70% (such as Figure 6 As shown). In terms of gas production stability, the system of the present invention uses the synergistic effect of the two-stage controller model to achieve the following in the early regulation stage (organic load rate is 1.0 - 1.5 g VS L -1 d -1 ), the gas production rate fluctuation range is less than ±5%, when the organic loading rate is increased to 2 g VS L -1 d -1 When the organic loading rate was 2 g VS L, the fluctuation increased to ±10%, but it could be adjusted quickly and reached a stable level of ±3% in the 5th week. -1 d -1 The fluctuation range reached over ±15% and was difficult to stabilize. The fluctuation range in control group 2 stabilized at around ±8% after the organic loading rate was increased, demonstrating inferior adjustment performance to the system of the present invention. These comparative experiments fully demonstrate the significant improvement in gas production and stability during anaerobic fermentation, as well as the advantages of the present invention's coordinated regulation of key parameters (such as organic loading rate and hydraulic retention time).

[0237] Based on the calculated K value range, divide it into equal proportions:

[0238] Calculate the boundary values of each interval

[0239] Table 1 Strategies adopted in different intervals

[0240] K value range pH changes Feeding strategy K>0.3 pH too low Increase feed rate −0.45≤K1≤0.3 Moderate pH Maintain normal feed K<−0.45 pH too high Reduce feed rate

[0241] If the pH exceeds 0.3, the feed rate is increased by 0.5 steps;

[0242] If the pH drops by less than 0.3 and increases by no more than 0.2, the feed rate remains unchanged;

[0243] If the pH increases by more than 0.2, the feed rate is reduced in steps of 0.1.

[0244] The phased changes and regulation obtained, such as Figure 3 As shown, the first to third weeks: organic loading rate is 1.0-1.5 gVS L -1 d -1 , gas production changes little, microorganisms completely consume organic matter, and the system operates stably. Starting from the fourth week: the organic loading rate is increased to 2 g VS L -1 d -1 After that, system fluctuations appear. At this time, by extending the material residence time or slowing down the feed rate, the system can gradually return to stability.

[0245] Combine Figure 2 and Figure 3 As shown, the experimental device of the present invention achieves the following control effects through the monitoring and adjustment functions of the data acquisition and control unit:

[0246] 1. Gas production rate stability:

[0247] (1) In the early stage of regulation (organic load rate is 1.0-1.5 g VS L -1 d -1 ), the gas production rate fluctuation range is within ±5%, which shows that the initial microorganisms have high efficiency in consuming organic matter and the system is stable; (2) when the organic loading rate is increased to 2 g VS L -1 d -1 When the system gas production rate fluctuated to ±10% for a time, by adjusting the feed rate and material residence time, the change in gas production rate gradually stabilized and was controlled within the range of ±3% in the 5th week.

[0248] 2. Response effect of dynamic adjustment: (1) The system automatically triggers the feed rate adjustment mechanism by real-time monitoring of the changing trend of gas production rate to achieve dynamic management of organic load; (2) When the gas production rate is detected to be rising, the control unit increases the feed rate to improve the reaction efficiency; when the gas production rate decreases, the control unit dynamically reduces the feed rate to reduce the organic load rate to prevent system overload; (3) During the entire experimental period, the dynamic regulation of the feed rate enables the system to quickly adapt to load changes at different stages.

[0249] 3. Data output and analysis: (1) As the organic loading rate gradually increases, the total gas production of the system shows a stable growth trend; when the organic loading rate is 2 g VS L -1 d -1 When the total gas production reaches 3.5 gVS L -1 d -1 In the high-load stage, the gas production is close to the theoretical maximum gas production value, which verifies the high fermentation efficiency of the device; (2) From the gas production rate change curve, it can be seen that the control strategy can effectively cope with the fluctuation of gas production rate and keep the system in the high-efficiency operation range; in the high-load stage, although the gas production rate changes greatly, the overall gas production maintains an upward trend, indicating that the device has a good absorption and regulation ability to disturbances; (3) During the operation of the system, when the organic load rate is adjusted to 2, 3, and 4 g VS L -1 d -1 When the feed rate fluctuates, the corresponding gas production fluctuation rate gradually decreases, the feed rate tends to be stable, and is effectively maintained in the high-efficiency operation range.

[0250] This example utilizes the platform's regulatory mechanisms to optimize the gas production process, from material pretreatment to real-time control. Combined with the experimental device's high-precision monitoring and dynamic regulation strategies, it can address gas production fluctuations at different stages, ensuring efficient and stable system operation.

[0251] Gas flow rate: The real-time measurement of gas production rate. Temperature: The temperature inside the reactor, which provides a basis for calculating the standardized gas flow rate and organic loading rate. Pressure: The pressure of the gas, which affects the volume measurement results.

[0252] During the fermentation process, special circumstances such as the absence of feed or inconsistent feed intervals can have the following impacts on the equipment: Failure to add feed can reduce the system's organic loading rate (OLR), thereby impacting microbial activity and reducing biogas production. Prolonged periods without feed addition can also lead to a decrease in microbial activity, affecting the system's overall fermentation efficiency. When no feed is added, the system adaptively adjusts operating conditions (such as reducing the agitation speed) to maintain a stable reaction environment.

[0253] Inconsistent dosing intervals can lead to fluctuations in the organic loading rate, directly affecting biogas production and causing periodic fluctuations in gas production, which is not conducive to the standardization and measurement of gas flow. Intermittent switching between high and low loads can affect the temperature, pH value, and microbial activity in the reactor, making the fermentation environment unstable.

[0254] The technical features of the above-mentioned embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above-mentioned embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0255] The above-described embodiments merely illustrate several implementations of the present invention, and while their descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the patent. It should be noted that a person skilled in the art would be able to make numerous variations and improvements without departing from the spirit of the present invention, all of which fall within the scope of protection of the present invention. Therefore, the scope of protection of the patent for this invention shall be determined by the appended claims.

Claims

1. An adaptive collaborative control method for modular CSTR fermentation simulation equipment, characterized in that: The steps include: (1) The gas production volume is measured in real time by the gas flow metering unit, the pH value of the reaction system is monitored by a pH meter or an online alkalinity meter, and the component characteristics of the produced gas are analyzed by the gas component monitoring unit; all monitoring data are collected and integrated by the data acquisition unit; (2) Based on the data collection results, the key process parameters of the system organic load are dynamically adjusted through the inlet and outlet control unit to meet the set operating goals; (3) The control system is constructed based on a multi-level control model, which includes an upper-level controller, a lower-level controller, an anaerobic process dynamic module, and an extreme value search model. The feed control unit adopts a two-level control architecture, in which the upper-level controller and the lower-level controller work together to achieve fine control of the reaction process. The expert decision feedback module is combined with the rule-driven extreme value search model to achieve dynamic optimization. Extreme value search strategy: When the pH setpoint reaches the lower limit, the system automatically reduces the feed rate to prevent a continued drop in pH. At the same time, it dynamically balances gas production efficiency and system stability by adjusting the upper and lower limits of the gas production rate, e2_max and e2_min. During the startup phase, an upper limit, e2_max, is set to accelerate microbial community enrichment and increase the initial gas production rate. When approaching the maximum system load, e2_max is lowered and the e2_min range is narrowed, reducing the feed step size and avoiding VFA accumulation. Dynamic threshold adjustment mechanism: By dynamically updating the upper and lower limits of the gas production rate difference, e2_max and e2_min, the gas production efficiency and system stability are balanced. Based on the coordinated response of the real-time gas production rate and pH, the core logic is as follows: ， Where, β=0.002(mg / L) -1 , which means that for every 1 mg / L increase in VFA concentration, e2_min increases by 0.002, pushing the system to increase feed to alleviate VFA accumulation; γ=0.15, which means that for every 1 unit deviation of pH from the set value, e2_min increases by 0.15, correcting acid-base imbalance by adjusting the feed rate; Transient balance control strategy: Dynamic control methods are designed for anaerobic fermentation system startup, load mutation, and external disturbance recovery. When the system is in a high-load transient state, the lower limit of the pH set value is maintained, and the following strategies are used to balance gas production and stability: ， Where λ is the attenuation coefficient. By real-time monitoring of pH, VFA, and gas production rate, when VFA>2000 mg / L, the feed rate is forced to be reduced to 50% of the baseline value; The lower-level proportional gain K1 in the lower-level controller is based on the current gas production rate Gas production rate at the previous moment The feed rate F(t) can be adjusted by adaptively changing the difference D(t) between them. ， The specific strategies of the expert database decision model are as follows: (1) When hour, The output value , at this time you can increase the feed; (2) When hour, , at this time, feed according to the original feed amount; (3) When When the last monitoring cycle hour, At this time, the feed should be stopped and the gas production changes should be observed for one day; in, This is the experience value after system debugging.

2. The method according to claim 1, characterized in that The method includes: Two-layer cascade control system, including: lower-level controller and upper-level controller; The lower controller is used to maintain the pH value within a set range; Control output: adjust feed speed; Proportional control: u1(t) = u 10 + K1× e1(t); Among them, u1(t): feed speed adjustment value, u 10 : Initial bias value of feed rate, K1: Lower stage proportional gain, e1(t): Error between actual pH value and set value e1(t) = pH - pH a ; The upper controller is used to optimize the biogas production of the system; Control output: Dynamically adjust the pH set value of the lower controller; Proportional-integral control: u2(t) = u 20 + K2 × e2(t) + K 2i ×∫e2(t)dt; Where, u2(t): the set value adjustment of the lower pH, u 20 : Initial pH setting value, K2: Upper proportional gain, K 2i : Upper integral gain, e2(t): The error between the target value and the actual value of gas production e2(t) = Q - Q a , where Q is the target gas production, Q a is the actual gas production, The lower-level controller uses a variable proportional gain PID control algorithm to adjust the feed flow in real time to maintain stable pH within the set range. The proportional gain KP of the PID controller is dynamically adjusted according to the change of ΔpH. The upper controller, based on the gas production rate Q a The deviation e2(t) from the target gas production rate Q is dynamically adjusted to adjust the target pH and feed control strategy.

Images