Anaerobic fermentation system driven by pyrolysis waste heat and optimization control method

Abstract

The embodiment of the invention provides an anaerobic fermentation system driven by pyrolysis waste heat. The system mainly comprises a flue gas heat exchanger, a condensation heat exchanger, a cooling heat exchanger, a medium-temperature water tank, a high-temperature water tank, a warming heat exchanger, a heat preservation heat exchanger, an organic solid waste pyrolysis unit, an anaerobic fermentation unit and an LSTM neural network central control unit. The embodiment of the invention further provides an optimal control method of the anaerobic fermentation system driven by the pyrolysis waste heat. According to the technical scheme, low-grade waste heat generated by the pyrolysis system is recycled and used for heating and temperature maintaining of the anaerobic fermentation unit, then the anaerobic fermentation temperature is controlled through LSTM neural network prediction, and the excellent self-adaption and non-linear adjustment capacity of a neural network algorithm is fully played; the interference of time sequence, uncertainty and volatility of a pyrolysis system and an anaerobic fermentation unit on the control process is overcome, and rapid, accurate and stable control over the anaerobic fermentation temperature is achieved.

Description

Technical Field

[0001] This invention belongs to the field of biomass energy utilization technology, and specifically relates to an anaerobic fermentation system driven by pyrolysis waste heat. This invention also provides an optimized control method for the aforementioned anaerobic fermentation system driven by pyrolysis waste heat. Background Technology

[0002] Anaerobic fermentation refers to the process of obtaining methane-rich biogas by decomposing organic matter through microorganisms under suitable conditions. It is one of the most efficient ways to utilize biomass energy on a large scale, and is particularly suitable for the efficient utilization of low-calorific-value biomass such as poultry and livestock manure and wet straw.

[0003] Temperature is one of the key factors affecting biogas production during anaerobic fermentation. Depending on the temperature of the biogas digester, anaerobic fermentation is generally divided into ambient temperature fermentation (10~30℃), mesophilic fermentation (30~40℃), and thermophilic fermentation (50~60℃). Temperature fluctuations also significantly impact anaerobic fermentation efficiency. Typically, daily temperature fluctuations in anaerobic fermentation must be controlled within ±2℃; when temperature fluctuations reach ±3℃, the anaerobic fermentation rate begins to be inhibited; and when temperature changes reach ±5℃, anaerobic gas production decreases significantly. In the cold environment of northern regions, where air temperatures are low and there are large temperature differences between winter and summer, and between day and night, appropriate heating and insulation measures are needed to maintain stable and efficient gas production. Strict control of the anaerobic fermentation temperature is also crucial to prevent interference from external environmental factors. Therefore, developing effective heating and insulation technologies for anaerobic fermentation is a prerequisite for developing large and medium-sized biogas projects in cold northern regions and is key to the large-scale promotion and application of biogas projects in these areas.

[0004] Currently, maintaining anaerobic fermentation temperatures in cold regions faces the problem of high energy consumption. Some scholars have proposed using renewable energy sources such as solar or geothermal energy for heating; however, the use of renewable energy is severely affected by geographical location, and the heating supply from renewable sources is unstable, necessitating the use of other energy sources for auxiliary heating. This results in a complex anaerobic fermentation temperature control system, hindering its widespread application. Furthermore, current anaerobic fermentation temperature control mainly employs the traditional PID algorithm. This algorithm often exhibits shortcomings such as large overshoot and long settling times when applied to systems with instability and time-varying characteristics, making it difficult to meet the high temperature control accuracy requirements of anaerobic fermentation. Therefore, there is an urgent need to develop a temperature control process with significant energy-saving benefits suitable for anaerobic fermentation in cold regions. Summary of the Invention

[0005] To overcome the shortcomings of the aforementioned prior art, this invention provides an anaerobic fermentation unit driven by pyrolysis waste heat utilization and its control method. By recovering the low-grade waste heat generated by the pyrolysis system and using it for heating and temperature maintenance of the anaerobic fermentation unit, and then using a Long Short-Term Memory (LSTM) neural network to predict and control the anaerobic fermentation temperature, a rapid, accurate and stable control of the anaerobic fermentation temperature can be achieved.

[0006] To address the aforementioned technical problems, the first aspect of this invention provides an anaerobic fermentation system driven by waste heat from pyrolysis, comprising an anaerobic fermentation unit, an organic solid waste pyrolysis unit, a medium-temperature water tank, a high-temperature water tank, a cooling heat exchanger, a condensing heat exchanger, a flue gas heat exchanger, a heating heat exchanger, a heat-insulating heat exchanger, and an LSTM neural network central control unit; wherein:

[0007] The anaerobic fermentation unit is used for anaerobic fermentation of raw materials; the organic solid waste pyrolysis unit is used for pyrolysis of organic solid waste to produce high-temperature pyrolysis char, pyrolysis gas and flue gas.

[0008] The cooling heat exchanger is connected to the organic solid waste pyrolysis unit, absorbs the heat of the pyrolysis carbon produced by pyrolysis, and cools the pyrolysis carbon at the same time. Its cooling water outlet is connected to the cooling water inlet of the condensing heat exchanger and the inlet of the medium temperature water tank. Part of the hot cooling water that has been heated after exchanging heat with the pyrolysis carbon is sent to the condensing heat exchanger and part is sent to the medium temperature water tank.

[0009] The condenser heat exchanger is connected to the organic solid waste pyrolysis unit, absorbs the waste heat of the pyrolysis gas generated by pyrolysis, and can condense its condensable components into pyrolysis oil for recovery. Its cooling water inlet is connected to the cooling water outlet of the cooling heat exchanger, and its cooling water outlet is connected to the inlet of the medium temperature water tank, so that the cooling water that has been heated further after exchanging heat with the pyrolysis gas is sent into the medium temperature water tank.

[0010] The flue gas heat exchanger is connected to the organic solid waste pyrolysis unit, absorbs the waste heat of the flue gas generated by pyrolysis, reduces the exhaust temperature, and improves the system thermal efficiency. Its cooling water outlet is connected to the inlet of the high-temperature water tank, and sends the hot water formed after the cooling water absorbs the waste heat of the flue gas into the high-temperature water tank.

[0011] The outlet of the medium-temperature water tank is connected to the process water inlet of the heating heat exchanger, and the process water outlet of the heating heat exchanger is connected to the cooling water inlet of the cooling heat exchanger. The fermentation broth of the anaerobic fermentation unit is input into the heating heat exchanger, and the fermentation broth that has heated up after exchanging heat with the process water is circulated back to the anaerobic fermentation unit to provide heat for anaerobic fermentation. The outlet of the high-temperature water tank is connected to the process water inlet of the heat-insulating heat exchanger, and the process water outlet of the heat-insulating heat exchanger is connected to the cooling water inlet of the flue gas heat exchanger. The fermentation broth of the anaerobic fermentation unit is input into the heat-insulating heat exchanger, and the fermentation broth that has heated up after exchanging heat with the process water is circulated back to the anaerobic fermentation unit to provide heat for anaerobic fermentation.

[0012] The LSTM neural network central control unit is connected to the organic solid waste pyrolysis unit, the medium-temperature water tank, the high-temperature water tank, and the anaerobic fermentation unit to collect control data. It then generates temperature control parameters based on the LSTM neural network model and sends the temperature control parameters to the heating heat exchanger and the heat preservation heat exchanger respectively. The heat output of the heating heat exchanger and the heat preservation heat exchanger is controlled according to the difference between the set temperature and the measured temperature of anaerobic fermentation, so that the measured temperature is consistent with the set temperature.

[0013] Preferably, the condenser heat exchanger is a liquid-liquid shell-and-tube heat exchanger.

[0014] Preferably, the cooling heat exchanger is a water-cooled spiral heat exchanger.

[0015] Preferably, the flue gas heat exchanger is a gas-liquid shell-and-tube heat exchanger.

[0016] In a second aspect, the present invention also provides an optimized control method for the aforementioned anaerobic fermentation system driven by waste heat from pyrolysis, the specific steps of which include:

[0017] 1. Time-series data are collected from the organic solid waste pyrolysis unit, the medium-temperature water tank, the high-temperature water tank, and the anaerobic fermentation unit;

[0018] 2. Perform data preprocessing on the system's historical operating data and collected time-series data;

[0019] 3. Create an LSTM neural network by dividing the preprocessed data into training and testing sets, and then using the training and testing sets to train the LSTM neural network model.

[0020] 4. Define the objective optimization function and use an LSTM neural network model to predict the future multi-step output of the anaerobic fermentation temperature under the current input control;

[0021] 5. Using an optimization algorithm, calculate the control input vector that minimizes the objective function, and use it as the current optimal control input vector;

[0022] 6. Adjust the temperature control parameters using the first component of the optimal control input vector, and output the temperature control parameters to regulate the anaerobic fermentation temperature;

[0023] 7. Repeat steps 1 to 6 to perform the next round of optimization control.

[0024] Preferably, in step 1, the time-series data consists of temperature data collected at preset time intervals and preset durations, including pyrolysis char temperature, pyrolysis gas temperature, flue gas temperature, medium-temperature water tank temperature, high-temperature water tank temperature, and anaerobic fermentation temperature. More preferably, the preset time interval is set to 1 min to 2 h, and the preset duration is not less than one month.

[0025] Preferably, in step 2, the data preprocessing includes data cleaning, data updating, and data normalization / standardization. More preferably, data cleaning involves removing invalid data; data updating involves extracting historical operational data with a duration longer than a preset duration, removing the earliest data within the preset duration, and adding newly collected time-series data of the preset duration to form the updated data; and data normalization / standardization involves normalizing / standardizing the updated data.

[0026] As a preferred embodiment of the above technical solution, in step 3, the data ratio of the training set and the test set is set to 7:3~8:2.

[0027] This invention employs a waste heat-driven anaerobic fermentation design. The system includes a flue gas heat exchanger, a condenser heat exchanger, and a cooling heat exchanger to recover waste heat from the high-temperature flue gas, pyrolysis gas, and pyrolysis char generated during the pyrolysis of organic solid waste. Medium-temperature and high-temperature water tanks are used as intermediate equipment to store the recovered waste heat. The fermentation broth in the anaerobic fermentation unit is then heated by connecting a water tank heating heat exchanger and a heat preservation heat exchanger, achieving temperature increase and insulation during the anaerobic fermentation process. A trained LSTM neural network model is used to automatically control the above waste heat recovery and utilization process. LSTM neural networks are currently the most widely used neural network algorithms, possessing powerful nonlinear fitting and adaptive energy adjustment capabilities, making them particularly suitable for time-series data analysis and processing.

[0028] The beneficial effects of the above-mentioned technical solution of the present invention are as follows:

[0029] 1. The flue gas, high-temperature pyrolysis gas, and pyrolysis char of the pyrolysis system have abundant low-grade waste heat. By recovering and utilizing the energy, the waste heat of pyrolysis can be recovered for heating of the anaerobic fermentation unit, which can avoid burning biogas or other methods to provide heat to maintain the anaerobic fermentation temperature, and has significant energy-saving benefits.

[0030] 2. Taking into full account the differences in temperature and energy quality of the flue gas, pyrolysis gas and pyrolysis char produced by the pyrolysis of organic solid waste, medium-temperature water tanks and high-temperature water tanks are used to store heat, which is then used for heating and insulation of anaerobic fermentation. This helps to overcome system fluctuations and achieve optimal energy matching and utilization.

[0031] 3. Based on the strong time-series characteristics of organic solid waste pyrolysis and anaerobic fermentation units, an LSTM neural network was created to perform nonlinear fitting and dynamic identification of anaerobic fermentation. Based on the trained LSTM neural network model, an optimal control method for the system within the predicted time was obtained, and only the first step of optimal control was implemented. This overcame the problem that the time-varying nature of the system would lead to a decrease in the prediction accuracy of the neural network, and significantly improved the accuracy, precision and stability of the system control. Attached Figure Description

[0032] The following figures are provided to further illustrate the invention and form part of the specification. They are used together with the detailed embodiments to explain the invention, but do not constitute a limitation thereof. They include:

[0033] Figure 1 This is a system connection diagram of an anaerobic fermentation system driven by waste heat from pyrolysis, provided in an embodiment of the present invention.

[0034] Figure 2 The flowchart illustrates the steps of an optimized control method for an anaerobic fermentation system driven by waste heat from pyrolysis, as provided in an embodiment of the present invention.

[0035] [Explanation of Key Component Symbols]

[0036] 1-Organic solid waste pyrolysis unit; 2-Anaerobic fermentation unit; 3-Medium temperature water tank; 4-High temperature water tank;

[0037] 131-Cooling heat exchanger; 132-Condensing heat exchanger; 141-Flue gas heat exchanger; 321-Heating heat exchanger; 421-Insulating heat exchanger;

[0038] 5-LSTM neural network control unit. Detailed Implementation

[0039] To make the technical problems, technical solutions, and advantages of the present invention clearer, a detailed description will be provided below in conjunction with the accompanying drawings and specific embodiments. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0040] To address the challenges of complex control systems for renewable energy heating in anaerobic fermentation, which hinders widespread adoption, and the shortcomings of traditional anaerobic fermentation temperature control algorithms (such as large overshoot and long settling times in unstable and time-varying systems), which fail to meet the temperature control precision requirements of anaerobic fermentation, a pyrolysis waste heat-driven anaerobic fermentation process was developed. This process recovers low-grade waste heat from the pyrolysis system and uses it for heating and temperature maintenance of the anaerobic fermentation unit. An LSTM neural network model is used to predict and control the anaerobic fermentation temperature, achieving rapid, accurate, and stable temperature control.

[0041] To achieve the above technical effects, such as Figure 1 As shown, this embodiment of the invention provides an anaerobic fermentation system driven by waste heat from pyrolysis, the components of which include:

[0042] Anaerobic fermentation unit 2, organic solid waste pyrolysis unit 1, medium-temperature water tank 3, high-temperature water tank 4, cooling heat exchanger 131, condensing heat exchanger 132, flue gas heat exchanger 141, heating heat exchanger 321, heat preservation heat exchanger 421, and LSTM neural network central control unit 5; wherein:

[0043] Anaerobic fermentation unit 2 is used for anaerobic fermentation of raw materials; organic solid waste pyrolysis unit 1 is used for pyrolysis of organic solid waste and to generate high-temperature pyrolysis char, pyrolysis gas and flue gas.

[0044] Cooling heat exchanger 131, condensing heat exchanger 132 and flue gas heat exchanger 141 are each connected to organic solid waste pyrolysis unit 1, and exchange heat by cooling pyrolysis carbon, condensing pyrolysis gas and absorbing waste heat from flue gas, respectively; one cooling water outlet of cooling heat exchanger 131 is connected to the cooling water inlet of condensing heat exchanger 132, and the other is connected to the inlet of medium temperature water tank 3. The cooling water outlet of condensing heat exchanger 132 is also connected to the inlet of medium temperature water tank 3. The cooling water outlet of flue gas heat exchanger 141 is connected to the inlet of high temperature water tank 4.

[0045] The outlet of the medium-temperature water tank 3 is connected to the process water inlet of the heating heat exchanger 321, and the process water outlet of the heating heat exchanger 321 is connected to the cooling water inlet of the cooling heat exchanger 131; the fermentation liquid of the anaerobic fermentation unit 2 is input into the heating heat exchanger 321, and the fermentation liquid that has been heated after exchanging heat with the process water is circulated back to the anaerobic fermentation unit 2 to provide heat for anaerobic fermentation.

[0046] The outlet of the high-temperature water tank 4 is connected to the process water inlet of the heat exchanger 421, and the process water outlet of the heat exchanger 421 is connected to the cooling water inlet of the flue gas heat exchanger 141. The fermentation liquid of the anaerobic fermentation unit 2 is input into the heat exchanger 421, and the fermentation liquid that has been heated after exchanging heat with the process water is circulated back to the anaerobic fermentation unit 2 to provide heat for anaerobic fermentation.

[0047] The LSTM neural network central control unit 5 is connected to the organic solid waste pyrolysis unit 1, the medium temperature water tank 3, the high temperature water tank 4 and the anaerobic fermentation unit 2 to collect control data. Then, it generates temperature control parameters based on the LSTM neural network model and sends the temperature control parameters to the heating heat exchanger 321 and the heat preservation heat exchanger 421 to regulate the heat output.

[0048] The fermentation broth circulated out of anaerobic fermentation unit 2 absorbs heat through heating heat exchanger 321 and heat preservation heat exchanger 421. The fermentation broth after heat absorption is used to maintain the anaerobic fermentation temperature in the anaerobic fermentation unit.

[0049] As a preferred embodiment, the condenser heat exchanger 132 may be a liquid-liquid shell-and-tube heat exchanger; the cooling heat exchanger 131 may be a water-cooled spiral heat exchanger; and the flue gas heat exchanger 141 may be a gas-liquid shell-and-tube heat exchanger.

[0050] This invention also provides an optimized control method for the above-mentioned anaerobic fermentation system driven by pyrolysis waste heat, such as... Figure 2 As shown, it includes the following steps:

[0051] S1: Collect time-series data from the organic solid waste pyrolysis unit, the medium-temperature water tank, the high-temperature water tank, and the anaerobic fermentation unit;

[0052] S2: Perform data preprocessing on the system's historical operating data and the collected time-series data;

[0053] S3: Create an LSTM neural network by dividing the preprocessed data into a training set and a test set, and using the training set and test set to train and obtain the LSTM neural network model.

[0054] S4: Define the objective optimization function and use the LSTM neural network model to predict the future multi-step output of the anaerobic fermentation temperature under the current input control;

[0055] S5: Using an optimization algorithm, calculate the control input vector that minimizes the objective optimization function, and use it as the current optimal control input vector;

[0056] S6: Adjust the temperature control parameters using the first component of the optimal control input vector, and output the temperature control parameters to regulate the anaerobic fermentation temperature;

[0057] S7: Repeat steps S1 to S6 to perform the next round of optimization control.

[0058] As a better implementation method:

[0059] In S1, the time-series data consists of temperature data collected at preset time intervals and preset durations. The temperature data includes the temperature of pyrolysis char, pyrolysis gas, flue gas, medium-temperature water tank, high-temperature water tank, and anaerobic fermentation. The preset time interval can be set from 1 min to 2 h, for example, 1 min, 5 min, 10 min, 30 min, 45 min, 1 h, 1.2 h, 1.5 h, and 2 h. The preset duration is set to be no less than 1 month, for example, 1 month, 3 months, 6 months, 10 months, 12 months, etc.

[0060] In S2, data preprocessing includes data cleaning, data updating, and data normalization / standardization. As a more specific implementation, the specific steps are as follows:

[0061] Data cleaning, also known as data cleaning, is the removal of invalid data.

[0062] Data update involves retrieving historical running data with a duration longer than the preset duration, removing the earliest data from the period with the preset duration, and adding newly collected time-series data of the preset duration to form the updated data.

[0063] Data normalization / standardization refers to the normalization / standardization process performed on updated data.

[0064] In S3, the ratio of training set to test set data is 7:3 to 8:2, for example, it can be divided into 7:3, 7.2:2.8, 7.5:2.5, 7.7:2.3 or 8:2.

[0065] As an example of the above-mentioned optimized control method, the preset time interval for collecting time-series data is 30 seconds, and the preset duration is 3 months. When updating the data, 12 months of historical operating data are retrieved, the earliest 3 months of data in the historical data are removed, and the newly collected 3 months of data are added to complete the data update. The ratio of training set to test set is 7:3. Under this optimized control method, the anaerobic fermentation temperature fluctuation during system operation does not exceed ±0.3℃, and the adjustment time does not exceed 5 minutes.

[0066] As another example of the above-mentioned optimized control method, the preset time interval for collecting time-series data is 1 minute, and the preset duration is 1 month. When updating data, 6 months of historical operating data are retrieved, the earliest data from the historical data is removed, and the newly collected data from the previous month is added to complete the data update. The ratio of training set to test set is 7:3. Under this optimized control method, the anaerobic fermentation temperature fluctuation during system operation does not exceed ±0.5℃, and the adjustment time does not exceed 10 minutes.

[0067] The technical solution of the above embodiment recovers the low-grade waste heat generated by the pyrolysis system and uses it for heating and temperature maintenance of the anaerobic fermentation unit. Then, the anaerobic fermentation temperature is controlled by LSTM neural network prediction. This fully utilizes the excellent adaptive and nonlinear adjustment capabilities of the neural network algorithm, overcomes the interference of the timing, uncertainty and fluctuation of the pyrolysis system and anaerobic fermentation unit on the control process, and realizes rapid, accurate and stable control of the anaerobic fermentation temperature.

[0068] In the description of this invention, unless otherwise expressly specified and limited, the terms "connection," "linked," and "connected" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Furthermore, the term "optimal" is used for descriptive purposes only and should not be construed as indicating or implying relative importance. Those skilled in the art will understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0069] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Units described as separate components may or may not be physically separate; components shown as units may or may not be physical units, i.e., they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0070] For the preferred embodiments of the present invention described above, common knowledge such as specific structures and characteristics in the technical solutions are not described in detail; each embodiment is described in a progressive manner, and the technical features involved in each embodiment can be combined with each other without conflicting with each other. The same or similar parts between the embodiments can be referred to each other.

[0071] It should be noted that, for those skilled in the art, the above embodiments can be improved and modified in various ways without departing from the principles of the present invention, and such improvements and modifications should also be considered to fall within the protection scope of the present invention.

Claims

1. An anaerobic fermentation system driven by waste heat from pyrolysis, comprising an anaerobic fermentation unit (2) for carrying out anaerobic fermentation, characterized in that, Also includes: The system comprises an organic solid waste pyrolysis unit (1), a medium-temperature water tank (3), a high-temperature water tank (4), a cooling heat exchanger (131), a condensing heat exchanger (132), a flue gas heat exchanger (141), a heating heat exchanger (321), an insulation heat exchanger (421), and an LSTM neural network central control unit (5); among which: The organic solid waste pyrolysis unit (1) is used to pyrolyze organic solid waste and generate high-temperature pyrolysis char, pyrolysis gas and flue gas; The cooling heat exchanger (131), condensing heat exchanger (132) and flue gas heat exchanger (141) are each connected to the organic solid waste pyrolysis unit (1) and exchange heat by cooling the pyrolysis carbon, condensing the pyrolysis gas and absorbing the waste heat of the flue gas, respectively. The cooling water outlet of the cooling heat exchanger (131) is connected to the cooling water inlet of the condensing heat exchanger (132) in one way and to the inlet of the medium temperature water tank (3) in the other way. The cooling water outlet of the condensing heat exchanger (132) is also connected to the inlet of the medium temperature water tank (3). The cooling water outlet of the flue gas heat exchanger (141) is connected to the inlet of the high temperature water tank (4). The outlet of the medium-temperature water tank (3) is connected to the process water inlet of the heating heat exchanger (321), and the process water outlet of the heating heat exchanger (321) is connected to the cooling water inlet of the cooling heat exchanger (131); the fermentation liquid of the anaerobic fermentation unit (2) is input into the heating heat exchanger (321), and the fermentation liquid that has been heated after exchanging heat with the process water is circulated back to the anaerobic fermentation unit (2) to provide heat for anaerobic fermentation; The outlet of the high-temperature water tank (4) is connected to the process water inlet of the heat exchanger (421), and the process water outlet of the heat exchanger (421) is connected to the cooling water inlet of the flue gas heat exchanger (141). The fermentation liquid of the anaerobic fermentation unit (2) is input into the heat exchanger (421), and the fermentation liquid that has been heated after exchanging heat with the process water is circulated back to the anaerobic fermentation unit (2) to provide heat for anaerobic fermentation. The LSTM neural network central control unit (5) is connected to the organic solid waste pyrolysis unit (1), the medium temperature water tank (3), the high temperature water tank (4) and the anaerobic fermentation unit (2) to collect control data, and then generates temperature control parameters based on the LSTM neural network model, and sends the temperature control parameters to the heating heat exchanger (321) and the heat preservation heat exchanger (421) respectively to regulate the heat output.

2. The anaerobic fermentation system according to claim 1, characterized in that, The condenser heat exchanger (132) is a liquid-liquid shell-and-tube heat exchanger.

3. The anaerobic fermentation system according to claim 1, characterized in that, The cooling heat exchanger (131) is a water-cooled spiral heat exchanger.

4. The anaerobic fermentation system according to claim 1, characterized in that, The flue gas heat exchanger (141) is a gas-liquid shell-and-tube heat exchanger.

5. The optimized control method for an anaerobic fermentation unit driven by pyrolysis waste heat utilization according to any one of claims 1 to 4, characterized in that, Includes the following steps: Step 1: Collect time-series data from the organic solid waste pyrolysis unit, the medium-temperature water tank, the high-temperature water tank, and the anaerobic fermentation unit; Step 2: Perform data preprocessing on the system's historical operating data and the collected time-series data; Step 3: Create an LSTM neural network by dividing the preprocessed data into a training set and a test set, and then using the training set and test set to train the LSTM neural network model. Step 4: Define the objective optimization function and use the LSTM neural network model to predict the future multi-step output of the anaerobic fermentation temperature under the current input control; Step 5: Using an optimization algorithm, calculate the control input vector that minimizes the objective optimization function, and use it as the current optimal control input vector; Step 6: Adjust the temperature control parameters using the first component of the optimal control input vector, and output the temperature control parameters to regulate the anaerobic fermentation temperature; Step 7: Repeat steps 1 to 6 to perform the next round of optimization control.

6. The optimized control method according to claim 5, characterized in that, In step 1, the time-series data is temperature data collected at preset time intervals and preset durations. The temperature data includes pyrolysis char temperature, pyrolysis gas temperature, flue gas temperature, medium-temperature water tank temperature, high-temperature water tank temperature, and anaerobic fermentation temperature.

7. The optimized control method according to claim 8, characterized in that, The preset time interval is 1 minute to 2 hours, and the preset duration is not less than one month.

8. The optimized control method according to claim 5, characterized in that, In step 2, the data preprocessing includes data cleaning, data updating, and data normalization / standardization.

9. The optimized control method according to claim 8, characterized in that, The data preprocessing includes the following steps: Data cleaning refers to the removal of invalid data; The data update involves retrieving historical running data with a duration longer than a preset duration, removing the earliest data from the period with the preset duration, and adding newly collected time-series data of the preset duration to form the updated data. The data normalization / standardization refers to the normalization / standardization process performed on the updated data.

10. The optimization control method according to any one of claims 5 to 9, characterized in that, In step 3, the ratio of training set to test set data is 7:3 to 8:2.

Images