Carbon dioxide-starch biosynthesis intelligent regulation and control chip system based on multi-mode AI

Abstract

The invention discloses a carbon dioxide-starch biosynthesis intelligent regulation and control chip system based on multi-mode AI. The system comprises a photonic crystal biosensor array used for monitoring physical and chemical parameters such as a pH value, a dissolved oxygen concentration and a carbon dioxide concentration in a reaction process in real time; the 16 * NPU array is used for efficiently calculating and processing the data of the multi-modal sensor and predicting the reaction dynamic state in real time; the 4 * Cortex-M7 treatment core is used for controlling the reaction process in real time and adjusting reactor conditions; and a neural symbol system-based multi-modal reaction regulation and control algorithm which is used for optimizing reaction dynamics and realizing cooperative regulation of 2000 + reaction parameters. All parameters in the reaction process are optimized through real-time multi-mode data monitoring (including enzyme activity, reaction temperature, pH value and the like). The enzyme activity can be accurately adjusted, so that the purity of the final product is remarkably improved, and the purity of the starch can be stabilized at 99% or above, while the purity of the starch prepared by the traditional method is only 89.7%.

Description

Technical Field

[0001] The present invention relates to the interdisciplinary technical field of synthetic biology and edge computing, and specifically to a carbon dioxide-starch biosynthesis intelligent regulation chip system based on multimodal AI. Background Art

[0002] The current traditional method of biosynthesizing starch faces several technical bottlenecks, especially in terms of efficient and stable control of the accuracy, rate and energy consumption of biochemical reactions: Enzyme activity fluctuations and low product purity: In the existing 11-step starch synthesis method of the Chinese Academy of Sciences, the activity of the enzyme fluctuates greatly in different batches and conditions, resulting in a product purity of only 89.7%, which cannot meet the demand for high-purity starch in industrial production; Reaction regulation delay: The traditional PID control algorithm has a long feedback delay (up to 4.5 seconds), which cannot guarantee timely response when dealing with the nonlinear and dynamic characteristics of biochemical reactions, resulting in an unstable reaction process and affecting product quality; Excessive energy consumption: The energy consumption of industrial-grade reactors is 32% higher than the theoretical value. Especially in the face of large-scale industrial production, the low energy efficiency and high energy consumption of equipment seriously limit the economic feasibility of this technology;

[0003] Shortcomings and challenges of existing technologies: Reaction parameters are difficult to control in real time: Due to the complex dynamic characteristics of biological reactions, traditional control strategies cannot quickly adapt to changes in the reaction environment, resulting in unstable reaction conditions and affecting conversion rate and purity. Unsuitable for extreme environments: Traditional technologies have poor adaptability to space, high pressure, and other extreme environments. In these environments, reaction control systems are easily affected by temperature, pressure, or radiation, making it difficult to maintain stable reaction performance.

[0004] To this end, we propose a carbon dioxide-starch biosynthesis intelligent regulation chip system based on multimodal AI. Summary of the Invention

[0005] To achieve the above objectives, the present invention provides the following technical solution: a carbon dioxide-starch biosynthesis intelligent regulation chip system based on multimodal AI, the system comprising:

[0006] Photonic crystal biosensor array: used to monitor physical and chemical parameters such as pH, dissolved oxygen concentration, and carbon dioxide concentration in real time during the reaction process;

[0007] 16×NPU array for efficient computation and processing of multimodal sensor data and real-time prediction of reaction dynamics;

[0008] 4 x Cortex-M7 processing cores: used to control the reaction process in real time and adjust the reactor conditions;

[0009] A multimodal reaction control algorithm based on a neural symbolic system: used to optimize reaction dynamics and achieve coordinated regulation of 2000+ reaction parameters.

[0010] Preferably, the photonic crystal biosensor array achieves electromagnetic shielding through TSV silicon vias and has a detection accuracy of pH ± 0.01 and concentration ± 0.1 ppm.

[0011] Preferably, the NPU array provides a computing power of 2.8TOPS and uses a deep learning model to predict enzyme activity and optimize reaction parameters in real time.

[0012] Preferably, the 4×Cortex-M7 processing core has high-speed processing capability, supports real-time data acquisition and feedback adjustment, and the response time does not exceed 10ms.

[0013] Preferably, the system can operate stably in extreme environments, including space radiation environment and deep sea high pressure environment.

[0014] Preferably, the chip adopts a sandwich structure, including a Cortex-M7 core on the upper layer, a photonic crystal sensor array on the lower layer, and an NPU array on the middle layer.

[0015] Preferably, the multimodal response control algorithm includes a distributed optimization framework based on federated learning, which can perform collaborative computing between different computing nodes.

[0016] Compared with the existing technology, the present invention provides a multimodal AI-based carbon dioxide-starch biosynthesis intelligent regulation chip system with the following beneficial effects:

[0017] This multimodal AI-based CO2-starch biosynthesis intelligent control chip system optimizes various parameters during the reaction process through real-time multimodal data monitoring (including enzyme activity, reaction temperature, pH, etc.). By precisely adjusting enzyme activity, the purity of the final product is significantly improved, with starch purity consistently exceeding 99%, compared to only 89.7% using traditional methods.

[0018] 2. This multimodal AI-based CO2-starch biosynthesis intelligent control chip system improves CO2 conversion by optimizing the CO2 reduction reaction step within the CO2-starch biosynthesis system. Experiments have shown that the CO2 conversion rate can reach 92.3%, significantly higher than the approximately 80% achieved by traditional methods. By real-time monitoring of reaction conditions and optimizing control, this system significantly reduces energy consumption. Compared to existing industrial-grade reactors, the system of this invention improves energy efficiency by approximately 32%, effectively reducing production costs and energy dependence.

[0019] 3. This multimodal AI-based carbon dioxide-starch biosynthesis intelligent regulation chip system takes full account of extreme environments (such as space, deep sea, etc.) and adopts tantalum metal shielding layers and silicon carbide packaging materials, so that the system can work normally in space radiation and deep-sea high-pressure environments. Experimental data show that in a space environment, the system has minimal impact on radiation performance; in a deep-sea environment, the system can withstand a pressure of 100MPa and continue to operate stably. Compared with traditional agricultural starch production, this system greatly improves resource utilization efficiency. By reducing land occupation (1 ton of starch requires less than 0.002 acres of land), reducing water consumption (only 5L of water per ton of starch), and shortening the production cycle (from 4 months to 3 days), it has significant advantages in sustainability and environmental protection. DETAILED DESCRIPTION

[0020] The following is a clear and complete description of the technical solutions in the embodiments of the present invention. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts are within the scope of protection of the present invention.

[0021] Example

[0022] Example of a carbon dioxide-starch biosynthesis intelligent regulation chip system based on multimodal AI

[0023] A multimodal AI-based carbon dioxide-starch biosynthesis intelligent regulation chip system, comprising:

[0024] Photonic crystal biosensor array: used to monitor physical and chemical parameters such as pH, dissolved oxygen concentration, and carbon dioxide concentration in real time during the reaction process;

[0025] 16×NPU array for efficient computation and processing of multimodal sensor data and real-time prediction of reaction dynamics;

[0026] 4 x Cortex-M7 processing cores: used to control the reaction process in real time and adjust the reactor conditions;

[0027] A multimodal reaction control algorithm based on a neural symbolic system: used to optimize reaction dynamics and achieve coordinated regulation of 2000+ reaction parameters.

[0028] Specifically, the photonic crystal biosensor array achieves electromagnetic shielding through TSV silicon vias and has a detection accuracy of pH ± 0.01 and concentration ± 0.1 ppm.

[0029] Specifically, the NPU array provides 2.8TOPS of computing power and uses a deep learning model to predict enzyme activity and optimize reaction parameters in real time.

[0030] Specifically, the 4×Cortex-M7 processing core has high-speed processing capability, supports real-time data acquisition and feedback adjustment, and has a response time of no more than 10ms.

[0031] Specifically, the system can work stably in extreme environments, including space radiation environment and deep sea high pressure environment.

[0032] Specifically, the chip adopts a sandwich structure, including a Cortex-M7 core on the upper layer, a photonic crystal sensor array on the lower layer, and an NPU array on the middle layer.

[0033] Specifically, the multimodal response control algorithm includes a distributed optimization framework based on federated learning, which can perform collaborative computing between different computing nodes.

[0034] Through the above technical solution, in the present invention, various parameters in the reaction process are optimized by real-time multimodal data monitoring (including enzyme activity, reaction temperature, pH value, etc.). Due to the ability to accurately adjust enzyme activity, the purity of the final product is significantly improved, and the purity of starch can be stabilized at more than 99%, while the purity of the traditional method is only 89.7%. The conversion rate of CO2 is improved by optimizing the CO2 reduction reaction step through the carbon dioxide-starch biosynthesis system. Experiments show that the conversion rate of CO2 can reach 92.3%, which is much higher than the 80% or so of the traditional method. By real-time monitoring of reaction conditions and optimizing control, this system can significantly reduce energy consumption. Compared with existing industrial-grade reactors, the system energy efficiency of the present invention is improved by about 32%, which can effectively reduce production costs and reduce dependence on energy. By fully considering extreme environments (such as space, deep sea, etc.), by adopting tantalum metal shielding layers and silicon carbide packaging materials, the system can work normally under space radiation and deep-sea high-pressure environments. Experimental data shows that radiation in space minimally impacts the system's performance. Furthermore, in deep-sea environments, the system can withstand pressures of 100 MPa and maintain sustained, stable operation. Compared to traditional agricultural starch production, this system significantly improves resource efficiency. By reducing land use (less than 0.002 mu of land is required for 1 ton of starch), lowering water consumption (only 5 liters of water per ton of starch), and shortening production cycles (from 4 months to 3 days), it offers significant advantages in sustainability and environmental protection.

[0035] 1. System hardware design

[0036] The intelligent control chip system of the present invention adopts a sandwich structure, which specifically includes the following modules:

[0037] Sensing layer (lower layer):

[0038] It includes a photonic crystal sensor array that can monitor the physical and chemical parameters such as pH, dissolved oxygen concentration, carbon dioxide concentration, etc. during the reaction process in real time.

[0039] The photonic crystal sensor adopts a specific electromagnetic shielding design and uses TSV silicon through-hole technology to effectively isolate electromagnetic interference and ensure high-precision data acquisition.

[0040] Computing layer (middle layer):

[0041] Equipped with 16 NPUs (neural processing units), each with a computing power of 2.8TOPS, it can perform efficient real-time data processing, execute deep learning algorithms, predict changes in enzyme activity during the reaction, and calculate the optimal reaction parameters.

[0042] A hybrid AI algorithm is used, including CNN (convolutional neural network) for spatial feature extraction, LSTM (long short-term memory network) for processing time series data, combined with a neural symbolic system to predict and optimize the reaction process.

[0043] Control layer (upper layer):

[0044] It uses four Cortex-M7 processing cores, which are specifically used to control the parameters of the reactor in real time, including temperature, pH value, reaction gas concentration, etc., to ensure the stability of the reaction process.

[0045] 2. Core algorithm design

[0046] Dynamic Metabolic Flux Analysis Algorithm (D-MFA):

[0047] This system monitors the various substances produced during the reaction (such as CO2 concentration and reaction products) in real time and uses a dynamic metabolic flux analysis algorithm to optimize the reaction pathway in real time. This algorithm can modify the metabolic pathway based on real-time data to maximize reaction efficiency.

[0048] Enzyme conformation prediction algorithm:

[0049] Using an improved version of the AlphaFold2 model, the enzyme's conformation is predicted to optimize enzyme activity. Based on the changing reaction environment, the algorithm dynamically adjusts the enzyme's structural state to maximize reaction speed and conversion rate.

[0050] Federated learning optimization framework:

[0051] Using a distributed optimization framework avoids overloading a single computing node. Multiple computing nodes (e.g., multiple chips or servers) work together through a federated learning protocol to achieve more efficient global optimization.

[0052] 3. Adaptability to extreme environments

[0053] Space Environment:

[0054] High-precision radiation shielding materials (such as tantalum metal shielding) are designed, and EDAC calibration technology is used to ensure data reliability in radiation environments. The system can operate continuously and stably in the low-temperature and high-radiation space environment.

[0055] Deep sea environment:

[0056] The system uses silicon carbide packaging materials, enabling it to operate normally in high-pressure environments (up to 100 MPa). The packaging material is also corrosion-resistant, allowing for long-term stable operation in deep sea environments.

[0057] 4. Industrial Application

[0058] In its application at the Jiangsu pilot plant, the system has successfully conducted 67 consecutive days of production experiments, with product purity remaining stable at 99.2% ± 0.3%, and energy consumption reduced by 58% compared to conventional production processes. This high-efficiency, low-energy production model demonstrates the system's potential for large-scale industrial production. While the embodiments of the present invention have been shown and described, it will be apparent to those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A multimodal AI-based carbon dioxide-starch biosynthesis intelligent regulation chip system, comprising: Photonic crystal biosensor array: used to monitor physical and chemical parameters such as pH, dissolved oxygen concentration, and carbon dioxide concentration in real time during the reaction process; 16×NPU array for efficient computation and processing of multimodal sensor data and real-time prediction of reaction dynamics; 4 x Cortex-M7 processing cores: used to control the reaction process in real time and adjust the reactor conditions; A multimodal reaction control algorithm based on a neural symbolic system: used to optimize reaction dynamics and achieve coordinated regulation of 2000+ reaction parameters.

2. The multimodal AI-based carbon dioxide-starch biosynthesis intelligent control chip system according to claim 1, characterized in that: The photonic crystal biosensor array achieves electromagnetic shielding through TSV silicon vias and has a detection accuracy of pH ±0.01 and concentration ±0.1 ppm.

3. The multimodal AI-based carbon dioxide-starch biosynthesis intelligent control chip system according to claim 1, characterized in that: The NPU array provides 2.8TOPS of computing power and uses a deep learning model to predict enzyme activity and optimize reaction parameters in real time.

4. The multimodal AI-based carbon dioxide-starch biosynthesis intelligent control chip system according to claim 1, characterized in that: The 4×Cortex-M7 processing core has high-speed processing capability, supports real-time data acquisition and feedback adjustment, and has a response time of no more than 10ms.

5. The multimodal AI-based carbon dioxide-starch biosynthesis intelligent control chip system according to claim 1, characterized in that: The system can work stably in extreme environments, including space radiation environment and deep sea high pressure environment.

6. The multimodal AI-based carbon dioxide-starch biosynthesis intelligent control chip system according to claim 1, characterized in that: The chip adopts a sandwich structure, including a Cortex-M7 core on the upper layer, a photonic crystal sensor array on the lower layer, and an NPU array on the middle layer.

7. The multimodal AI-based carbon dioxide-starch biosynthesis intelligent control chip system according to claim 1, characterized in that: The multimodal response control algorithm includes a distributed optimization framework based on federated learning, which can perform collaborative computing between different computing nodes.