Prediction method for optimal switching time of microalgae culture stage

Abstract

The invention discloses a method for predicting the optimal switching time of a microalgae culture stage, relates to the crossing field of biotechnology and deep learning, and can quantify and predict the optimal switching time. A model construction experiment of about 14 days is completed on a small-scale system, then a tedious full-cycle optimization experiment does not need to be carried out on any scale-expanded system, and the optimal conversion time can be determined by monitoring the biomass concentration and the cell concentration and substituting the biomass concentration and the cell concentration into a prediction model for calculation. The optimization time is shortened from several weeks to several days, and a large amount of manpower, material resources and time cost are saved.

Description

Technical Field

[0001] This invention relates to the interdisciplinary field of biotechnology and deep learning, specifically to a method for predicting the optimal switching time during microalgae cultivation. Background Technology

[0002] Microalgae are a high-value biological resource, rich in lipids, proteins, carbohydrates, and various bioactive substances such as astaxanthin and polyunsaturated fatty acids. Compared with traditional crops, microalgae grow faster, have higher photosynthetic efficiency, stronger carbon fixation capacity, and can utilize nutrients and CO2 from wastewater and exhaust gases for production on non-arable land such as saline-alkali land, making them an environmentally friendly and sustainable raw material. However, the industrialization of microalgae still faces key challenges: on the one hand, the accumulation of target products (such as lipids and pigments) often depends on stress conditions (such as nitrogen deficiency and high light), which inhibits microalgae growth; on the other hand, large-scale cultivation results in low product content and high upstream and downstream costs, leading to insufficient economic viability. Therefore, how to synergistically improve microalgae biomass and product content is a core issue that urgently needs to be addressed to achieve its commercialization.

[0003] Two-stage microalgal culture is one of the mainstream strategies for increasing microbial lipid production. It involves rapidly accumulating biomass under optimal conditions in the first stage, followed by inducing massive lipid synthesis under stress conditions in the second stage. The timing of switching from the first stage to the second stage (switch-in time) is crucial in determining the final cost and benefit.

[0004] Currently, the most similar and widely used existing technology both domestically and internationally is the "biomass peak conversion method." This method involves continuously monitoring the biomass concentration during the first stage of microalgae cultivation. When the biomass concentration reaches its maximum, indicating the early stage of the growth curve's plateau, a stage switch is immediately initiated, harvesting the algal cells and transferring them to the second stage of cultivation under stress conditions. This method is considered a standard operating procedure and is frequently mentioned in numerous academic papers and patent descriptions. However, this existing technology has significant drawbacks: 1. Suboptimal yield: Recent studies have shown that peak biomass does not necessarily indicate that cells are in their optimal physiological state. At the time of peak biomass, the proliferative capacity of microalgal cells often begins to decline. Switching at this point, while initially resulting in high biomass, leads to insufficient subsequent growth capacity under the second stage of stress, resulting in a final lipid yield that is not optimal.

[0005] 2. Lack of predictability and inefficiency: This method is a "post-hoc" judgment, requiring the biomass to reach its peak before operation, making advance prediction and optimization impossible. For new algal species, new reactors, or new stress conditions, researchers must conduct numerous and time-consuming experiments (usually requiring 2-3 complete growth cycles) throughout the entire growth cycle to re-explore the optimal timing. This trial-and-error process is costly and extremely inefficient, severely hindering the industrial application of this technology. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a method for predicting the optimal switching time during microalgae cultivation. By establishing a model through small-scale experiments, this method can guide production systems of different scales, significantly reducing the time and resource costs required for optimization.

[0007] To achieve the above-mentioned objectives, the technical solution adopted by this invention is as follows: A method for predicting the optimal switching time during microalgae cultivation is provided, comprising the following steps: S1: The target microalgae were cultured in the first-stage culture medium under non-stress conditions. The biomass concentration and cell concentration of the microalgae were collected at fixed times every day during the mid-to-late logarithmic growth phase. At each measurement time, algal cells at the measurement time point were taken and cultured in the second-stage culture medium for 7 days. The lipid concentration of the microalgae taken each time after 7 days of culture in the second stage was measured. The maximum lipid concentration Y and its corresponding sampling time were determined. The sampling time corresponding to the maximum lipid concentration Y was used as the switching time, and the switching time was used as the label of the maximum lipid concentration. The standard sample group was then integrated. Using the same method, several different stress environments were set during the second stage of cultivation, and several stress sample groups under different stress environments were integrated to obtain several stress sample groups under different stress environments; The late logarithmic growth phase of microalgae culture is from day 8 to 14 of culture in the first-stage medium. S2: Based on the multiple linear regression model, a standard prediction model and several stress prediction models were established respectively; S3: Normalize the biomass concentration and cell concentration in the standard sample group and several stress sample groups and use them as training sets to train the standard prediction model and several corresponding stress prediction models, respectively, to obtain the trained standard prediction model and several stress prediction models. S4: During the first stage of microalgae culture, the biomass concentration and cell density are periodically measured, and environmental data of the microalgae culture are collected to determine whether the environment is a standard environment or a stress environment. The biomass concentration and cell density are then input into the trained prediction model for the corresponding environment to obtain the predicted maximum lipid concentration. When the prediction result of the latest xth monitoring is less than or equal to the prediction result of the x-1th monitoring, the time of the xth monitoring is the optimal switching time. The time interval between two monitoring is 1~24h.

[0008] Furthermore, the time interval between the two monitoring sessions was 3 hours.

[0009] Furthermore, stressful environments include nitrogen-deficient environments, nitrogen-rich environments, high light intensity environments, low light intensity environments, phosphorus-deficient environments, phosphorus-rich environments, carbon-deficient environments, carbon-rich environments, low temperature environments, high temperature environments, and high salinity environments.

[0010] Furthermore, when the stress environment is a nitrogen-deficient environment, the nitrogen-deficient stress prediction model is Y 缺氮 =0.501×Xs1+0.969×Xs2-0.376; where Xs1 is the normalized value of biomass concentration and Xs2 is the normalized value of cell concentration.

[0011] Furthermore, when the stress environment is a high light intensity environment, the high light intensity stress prediction model is Y. 高光强 =0.442×Xs1+0.805×Xs2-0.193; where Xs1 is the normalized value of biomass concentration and Xs2 is the normalized value of cell concentration.

[0012] Furthermore, the normalization is performed as min-max normalization, Xs=(X-Xmin) / (Xmax-Xmin); where X is the biomass concentration or cell concentration; Xs is the normalized value of the biomass concentration or cell concentration; Xmin is the minimum biomass concentration or minimum cell concentration in the sample group; and Xmax is the maximum biomass concentration or maximum cell concentration in the sample group.

[0013] Furthermore, it also includes conducting Pearson correlation analysis on data from several stress sample groups under different stress environments to determine the weight index corresponding to different stress environments. When multiple stresses occur simultaneously in the first-stage culture environment, the stress prediction model with the largest weight among the multiple stresses is selected for prediction.

[0014] The present invention also provides a non-transitory computer-readable storage medium storing at least one instruction or at least one program, wherein the at least one instruction or at least one program is loaded and executed by a processor to implement the method for predicting the optimal switching time of the microalgae cultivation stage as described above.

[0015] The present invention also provides an electronic device, including a processor and the aforementioned non-transitory computer-readable storage medium.

[0016] The beneficial effects of this invention are as follows: 1.) Precise prediction, significantly improving efficiency and reducing costs: The method of this invention can quantitatively predict the optimal conversion time. A model building experiment of approximately 14 days is completed on a small-scale system. Thereafter, for any scaled-up system, there is no need for lengthy full-cycle optimization experiments. The optimal conversion time can be determined simply by monitoring biomass and cell concentrations and inputting them into the predictive model. This reduces optimization time from weeks to days, saving significant human, material, and time costs.

[0017] 2.) High versatility and reliability: The method has been validated under two different stress conditions (nitrogen deficiency and high light intensity) and has successfully predicted and guided the production of 7L systems from 3L systems, demonstrating its effectiveness and reliability under different conditions and scales.

[0018] 3.) Easy to automate: It is easy to integrate with existing process sensors and automatic control systems, laying the foundation for intelligent cultivation and precise control of microalgae. Attached Figure Description

[0019] Figure 1 The examples show the prediction results under nitrogen deficiency and high light intensity stress; Figure 2 The results are actual measurements taken in a 7L reactor in the examples. Detailed Implementation

[0020] The specific embodiments of the present invention are described below to enable those skilled in the art to understand the present invention. However, it should be understood that the present invention is not limited to the scope of the specific embodiments. For those skilled in the art, various changes are obvious as long as they are within the spirit and scope of the present invention as defined and determined by the appended claims. All inventions utilizing the concept of the present invention are protected.

[0021] The embodiments of this invention use Chlorella pyrenoidosa. Auxenochlorellapyrenoidosa As the target microalgae.

[0022] Example 1 A method for predicting the optimal switching time in microalgae cultivation stages, comprising the following steps: S1: Target microalgae were inoculated into a 3L photobioreactor and cultured under optimal conditions in the first stage. Starting from the late logarithmic growth phase (day 8), samples were taken at fixed time intervals until the growth entered the apoptosis phase (day 14), specifically on days 8, 9, 10, 11, 12, 13, and 14. Seven sets of key cellular characteristic parameters were obtained, including: a) Biomass concentration (X1): determined by centrifugation and drying, in g / L. b) Cell density (X2): determined by a hemocytometer or automated cell counter, in cells / L.

[0023] Seven groups of microalgae were cultured for 7 days under nitrogen-deficient and high light intensity conditions, respectively. The lipid concentration of the microalgae after the second stage of culture was measured each time. The maximum lipid concentration Y was taken, and the measurement time point corresponding to the maximum lipid concentration Y was used as the switching time. The switching time was used as the label of the maximum lipid concentration. The samples were integrated to obtain the nitrogen-deficient stress sample group and the high light intensity stress sample group. Stressful environments can also include nitrogen-rich environments, low light intensity environments, phosphorus-deficient environments, phosphorus-rich environments, carbon-deficient environments, carbon-rich environments, low temperature environments, high temperature environments, and high salinity environments.

[0024] S2: Nitrogen deficiency stress prediction model and high light intensity stress prediction model were established based on the multiple linear regression model; S3: Normalize the biomass concentration and cell concentration (min-max) in the nitrogen deficiency stress sample group and the high light intensity stress sample group, Xs = (X - Xmin) / (Xmax - Xmin); where X is the biomass concentration or cell concentration; Xs is the normalized value of the biomass concentration or cell concentration; Xmin is the minimum biomass concentration or minimum cell concentration in the sample group; and Xmax is the maximum biomass concentration or maximum cell concentration in the sample group. Integrate the normalized values ​​and the corresponding maximum lipid concentration Y sets to obtain the nitrogen deficiency stress training set and the high light intensity stress training set, respectively. Train the nitrogen deficiency stress prediction model and the high light intensity stress prediction model, respectively, to obtain the trained nitrogen deficiency stress prediction model Y. 缺氮 =0.501×Xs1+0.969×Xs2-0.376; and the high light intensity stress prediction model Y 高光强 =0.442×Xs1+0.805×Xs2-0.193.

[0025] S4: The target microalgae, *Chlorella proteoglycans*, was cultured in a 7L photobioreactor under the same conditions. Starting from day 8, the biomass concentration and cell density were monitored daily. After normalizing the data, they were substituted into the prediction model obtained in S3 for prediction. The prediction results are as follows. Figure 1 As shown, by Figure 1 It is known that the lipid concentration is predicted to reach its peak on day 13. In practice, when the predicted result is equal to or less than the previous prediction, the culture phase should be switched immediately to ensure that the microalgae culture achieves optimal lipid quality. Biomass concentration and cell density change little within 1 hour, and an interval of less than 1 hour between two monitoring sessions would only waste resources, while an interval that is too long would lead to a significant decrease in the predicted maximum lipid concentration. Therefore, a 3-hour interval between two monitoring sessions is optimal, which does not increase monitoring pressure, while allowing for timely response and avoiding missing the peak of the maximum lipid concentration, thus achieving the optimal culture switching time.

[0026] To verify the accuracy of the prediction results, switching experiments were conducted on days 12, 13, 14, and 15. The lipid concentration (oil concentration) was measured on days 4-7 of the second-stage culture after the switching. The results are as follows: Figure 2 As shown, by Figure 2It can be seen that after the transition from the first stage to the second stage on the 13th day, the lipid concentration was at its maximum after 7 days of culture, regardless of whether the environment was nitrogen-deficient or high light intensity; the measured results are consistent with the prediction of the prediction model.

Claims

1. A method for predicting the optimal switching time in a microalgae cultivation stage, characterized in that, Includes the following steps: S1: The target microalgae were cultured in the first-stage culture medium under non-stress conditions. The biomass concentration and cell concentration of the microalgae were collected at fixed times every day during the mid-to-late logarithmic growth phase. At each measurement time, algal cells at the measurement time point were taken and cultured in the second-stage culture medium for 7 days. The lipid concentration of the microalgae taken each time after 7 days of culture in the second stage was measured. The maximum lipid concentration Y and its corresponding sampling time were determined. The sampling time corresponding to the maximum lipid concentration Y was used as the switching time, and the switching time was used as the label of the maximum lipid concentration. The standard sample group was then integrated. Using the same method, several different stress environments were set during the second stage of cultivation, and several stress sample groups under different stress environments were integrated to obtain several stress sample groups under different stress environments; The late logarithmic growth phase of the microalgae culture is from day 8 to 14 of culture in the first-stage culture medium; S2: Based on the multiple linear regression model, a standard prediction model and several stress prediction models were established respectively; S3: Normalize the biomass concentration and cell concentration in the standard sample group and several stress sample groups and use them as training sets to train the standard prediction model and several corresponding stress prediction models, respectively, to obtain the trained standard prediction model and several stress prediction models. S4: During the first stage of microalgae culture, the biomass concentration and cell density are periodically measured, and environmental data of the microalgae culture are collected to determine whether the environment is a standard environment or a stress environment. The biomass concentration and cell density are then input into the trained prediction model for the corresponding environment to obtain the predicted maximum lipid concentration. When the prediction result of the latest xth monitoring is less than or equal to the prediction result of the x-1th monitoring, the time of the xth monitoring is the optimal switching time. The time interval between two monitoring is 1~24h.

2. The method for predicting the optimal switching time in microalgae cultivation stages according to claim 1, characterized in that, The time interval between the two monitoring sessions was 3 hours.

3. The method for predicting the optimal switching time in microalgae cultivation stages according to claim 2, characterized in that, The stress environments include nitrogen-deficient environments, nitrogen-rich environments, high light intensity environments, low light intensity environments, phosphorus-deficient environments, phosphorus-rich environments, carbon-deficient environments, carbon-rich environments, low temperature environments, high temperature environments, and high salinity environments.

4. The method for predicting the optimal switching time in microalgae cultivation stages according to claim 3, characterized in that, When the stress environment is a nitrogen-deficient environment, the nitrogen-deficient stress prediction model is Y. 缺氮 =0.501×Xs1+0.969×Xs2-0.376; where Xs1 is the normalized value of biomass concentration and Xs2 is the normalized value of cell concentration.

5. The method for predicting the optimal switching time in microalgae cultivation stages according to claim 3, characterized in that, When the stress environment is a high light intensity environment, the high light intensity stress prediction model is Y. 高光强 =0.442×Xs1+0.805×Xs2-0.193; where Xs1 is the normalized value of biomass concentration and Xs2 is the normalized value of cell concentration.

6. The method for predicting the optimal switching time in microalgae cultivation stages according to claim 1, characterized in that, The normalization is min-max normalization, Xs=(X-Xmin) / (Xmax-Xmin); where X is the biomass concentration or cell concentration; Xs is the normalized value of the biomass concentration or cell concentration; Xmin is the minimum biomass concentration or minimum cell concentration in the sample group; and Xmax is the maximum biomass concentration or maximum cell concentration in the sample group.

7. The method for predicting the optimal switching time in microalgae cultivation stages according to claim 1, characterized in that, It also includes Pearson correlation analysis of data from several stress sample groups under different stress environments to determine the weight index corresponding to different stress environments. When multiple stresses occur simultaneously in the first-stage culture environment, the stress prediction model with the largest weight among the multiple stresses is selected for prediction.

8. A non-transitory computer-readable storage medium, wherein the storage medium stores at least one instruction or at least one program segment, characterized in that, The at least one instruction or the at least one program segment is loaded and executed by the processor to implement the method for predicting the optimal switching time of the microalgae cultivation stage as described in any one of claims 1 to 6.

9. An electronic device, characterized in that, Includes a processor and the non-transitory computer-readable storage medium as described in claim 8.

Images