CN106155123B - Online control method for residual sugar concentration in fermentation process - Google Patents

Abstract

The invention expands the expert knowledge method, adopts the fuzzy control method to simulate the expert knowledge method, and proposes a real-time control method of sugar concentration. This method takes the residual sugar concentration and its change rate in the tank as the input of the fuzzy control, and the sugar replenishment rate as the output. The input and output membership functions of the model controller, the fuzzy control rules of the model controller and the clarification algorithm are used to simulate fuzzy control based on expert experience. In online application, the computer controls the constant-flow pump to extract the fermentation broth and dilutes it, and then controls the YSI2700 online analyzer for analysis, and then substitutes the hysteresis detection value into the designed fuzzy controller to obtain the compensation in the control period. sugar rate. The method is the optimization and general expansion of the expert knowledge method, which can greatly improve the product quality and yield of the fermentation process.

Description

A kind of online control method of residual sugar concentration in fermentation process

technical field

The invention belongs to the field of application engineering, and particularly relates to an on-line control method of residual sugar concentration in a fermentation process using a hysteresis detection value.

Background technique

In the fermentation process, sugar is usually used as a nutrient for bacterial growth, and its concentration directly determines the quality and yield of the product. If the sugar concentration is too low, the bacteria are in a state of insufficient nutrition and grow slowly; while the sugar concentration is too high, the growth of the bacteria is also inhibited due to excess nutrition. Limited by the development of detection technology, the traditional method of offline sampling and analysis is mostly used for the detection of sugar concentration at present, which has a large detection period and a long lag time. The residual sugar concentration control based on this detection technology is relatively extensive, and the fine control of the fermentation process cannot be further realized. In addition, the traditional detection method involves manual participation, which is a bottleneck problem in realizing the comprehensive automation of the fermentation process. At present, the control methods of residual sugar concentration often use pH-star and DO-star. According to the relationship between sugar consumption and pH and DO during the growth of fermentation broth, the current online pH and DO detection values are used to control residual sugar concentration. , This method is an indirect control method, which sacrifices the growth rate of microorganisms to control the accumulation of metabolic by-products, so that the microorganisms are in a state of substrate scarcity for a long time, and due to the relatively large fluctuations in pH and DO, it cannot reach very high bacterial cells. concentration. YSI2700 can detect residual sugar concentration online, which provides the possibility for real-time online control of residual sugar concentration. However, the detection range of YSI2700 is limited, and it is easily affected by external factors such as temperature, stirring, foam, etc. The fermentation broth needs to be extracted for detection. However, the extraction process takes a long time, so there is a large lag in the detection value. Based on the detection lag value, the present invention designs a fuzzy-based control strategy based on expert experience and knowledge, which can be generally used in various fermentation processes, and solves the disadvantage that residual sugar concentration control relies on expert experience.

SUMMARY OF THE INVENTION

The purpose of the present invention is to overcome the shortcomings of the detection of residual sugar concentration in the existing fermentation process, and the shortcomings of the control strategy based on expert experience, such as relying on expert experience, narrow application and difficult to popularize and apply, provide a fuzzy control strategy, It is used to summarize the expert experience and deduce it into a general formula expression, which provides a new path for the control of the residual sugar concentration in the fermentation process.

In view of the large lag in the detection of sugar concentration by the YSI2700 online sugar analyzer, fermentation experts judge the sugar consumption rate of different growth stages of the bacteria according to their experience, and propose a residual sugar concentration control strategy, which is based on expert experience. control strategy. Aiming at the detection system, based on the hysteresis detection value of the YSI2700 instrument, the invention summarizes and deduces the expert control experience through the fuzzy control algorithm, and proposes a fuzzy control strategy for the residual sugar concentration in the fermentation process, so that the expert control experience can be used in the residual sugar in the fermentation process. Concentration control has been widely promoted.

To achieve the above object, the present invention comprises the steps:

In order to realize the on-line detection of residual sugar by YSI2700, the computer is used as the core controller, and the auxiliary constant current pump TBP1010 and TBP1002 are used. The computer is connected with the YSI2700 and the constant current pump through the serial port, respectively, and the communication is completed by the protocols of the YSI2700 and the constant current pump. Using this detection system, firstly, the computer controls the constant-flow pump to extract the fermentation broth, and at the same time controls the constant-flow pump to add water for dilution, and then controls the YSI2700 online sugar analyzer to detect the sugar concentration. The system is developed by VC and has functions such as real-time data display, manual automatic control of YSI2700 and constant-flow pump, and fuzzy control of sugar concentration by historical data curve display.

The detection lag time was calculated according to the extraction rate of the fermentation broth, the length of the pipeline and the analysis time of the YSI2700.

Taking the residual sugar concentration in the tank and its rate of change as the input of the fuzzy control, and the sugar replenishment rate as the output, by determining the input and output language variables of the model controller, the universe of discourse and the control method, quantification factor and scale factor, model control The input and output membership functions of the controller, the fuzzy control rules of the model controller and the clearing algorithm are used to simulate the fuzzy control based on the expert experience, and the fuzzy control parameters are adjusted online, so that the residual sugar control effect can be optimal;

The designed fuzzy control strategy is used in the fermentation process: first, the sugar concentration is analyzed by the detection system based on the YSI2700 online analyzer, and then the hysteresis detection value of the residual sugar concentration is substituted into the designed fuzzy controller to obtain the control. Glucose replenishment rate during the cycle.

Description of drawings

Fig. 1 is the flow chart of sugar supplementation control

Figure 2(1) is the flow chart of residual sugar concentration control based on expert experience knowledge method (the output detection value is less than the upper limit for the first time)

Figure 2(2) is a flow chart of residual sugar concentration control based on expert experience and knowledge method (initial value of sugar supplementation rate)

Figure 2(3) is the flow chart of residual sugar concentration control based on expert experience and knowledge method (normal sugar supplementation)

Figure 3(1) Flowchart of residual sugar concentration control based on fuzzy control algorithm (the output detection value is less than the upper limit for the first time)

Figure 3(2) Flowchart of residual sugar concentration control based on fuzzy control algorithm (initial value of sugar supplementation rate)

Figure 3(3) Flowchart of residual sugar concentration control based on fuzzy control algorithm (normal sugar supplementation)

Detailed ways

The structure of the present invention will be further described below with reference to FIG. 1 .

As shown in Figure 1, the thick solid lines are pipes, the thin solid lines are communication lines, and the arrows indicate the transmission direction of signals or liquids. The equipment involved in the present invention includes: constant flow pump 1 (TBP1010), constant flow pump 2 (TBP1002), on-line sugar concentration analyzer (YSI2700), fermenter, peristaltic pump, mixer (dilution), and PC. The constant flow pump 1 extracts the reaction liquid from the fermenter through the pipeline, and the constant flow pump 2 is connected to pure water. The two liquids are fully mixed and diluted in the mixer according to a certain ratio, and sent to the online sugar concentration analyzer for online detection. At the same time, The liquid in the mixer overflows, and the detected value is sent to the PC through the communication line, and is controlled by the algorithm to adjust the opening of the peristaltic pump and supply sugar to the fermenter.

The residual sugar concentration control process of the expert experience and knowledge method is further described in conjunction with FIG. 2 .

Set the flow rate of the constant flow pump and the sampling cycle of the online sugar analyzer, and run as follows: First, the system starts to run, and the initial value g=0, i=1 (g is used to judge whether the detected value Cn(i) is greater than The upper limit a, i represent the detection time). The function of the flowchart (1) is to output the time g when the detected value is smaller than a for the first time. In the flow chart (2), when i<=g, Cn(i)>a is always established, the sugar feeding rate is 0; once i>g occurs, the initial value of the sugar feeding rate F(i-1)=ΔF (ΔF is the fixed sugar supplementation rate), end the process. Flowchart (3) shows the situation of normal sugar supplementation at time i>g. It can be clearly seen from the flowchart (3) that the upper limit a, the expected value Cs, the lower limit b and the detection value Cn(i-1) at the previous moment need to be logically judged, respectively. When F(i)=F( In the case of i-1)-ΔF, it is necessary to judge F(i) and 0, because the sugar supplementation rate cannot be negative. In the case of the above different logical judgments, the relative sugar supplementation rate is obtained, from which we can know the value E(i) in the fermenter under different sugar supplementation rates, (E(i)=f[Fn(i)] , indicating the relationship between the sugar concentration in the fermenter and the sugar replenishment rate), and finally determine whether the detection time i is greater than m (m represents the number of times to be detected), if it is less than, the system will perform the next cycle, otherwise, end the operation of the system .

The fuzzy control flow chart of the residual sugar concentration is further described in conjunction with FIG. 3 .

Set the flow rate of the constant-flow pump and the sampling period of the online sugar analyzer, and run as follows: First, the system starts to run, and the functions of the flowcharts (1) and (2) are the same as those of the expert experience and knowledge method, and will not be repeated. , the only difference is that in the flow chart (2), when i>g occurs, the initial sugar supplementation rate is different. In the flow chart (3), when the detected value Cn(i)>the upper limit a, the sugar replenishment rate is 0 (because the detection time of YSI2700 is short, the sugar concentration in the fermenter cannot be consumed instantly). And because the sugar replenishment rate F(i) is related to the previous moment F(i-1), it must be counted, here is counted by n (assuming that the detected value at time i is Cn(i)<a, and Cn(i-1)> a, F(i-1)=0, F(i-1) cannot be used to calculate F(i), and the value at time F(i-1-n) must be used for calculation to ensure that F(i- 1-n)>0). In addition, when the detected value Cn(i) < lower limit b, it is necessary to judge the detected value Cn(i-1) at the previous moment, and add the sugar supplementation rate at the last moment or the last 1+n moment. Upper ΔF (ΔF is consistent with the fixed sugar supplementation rate of the expert experience method, of course, it can also be changed according to the site conditions). When the detected value a<Cn(i)<b, the fuzzy processing is performed first, and then the fuzzy controller is entered to output Δu (Δu is the sugar supplement rate value that needs to be changed). The internal rules of the fuzzy controller are to summarize the expert experience It is formulated on the basis of the knowledge method. Finally, it is also necessary to make a logical judgment on F(i) and 0. The next operation steps are the same as the expert experience knowledge method.

Claims (1)

1. a fuzzy control method based on hysteresis residual sugar concentration detection value, is characterized in that, comprises the steps: (1) Construction of an online control system for the concentration of residual sugar in the fermentation process: the computer is used as the core controller, and the auxiliary constant-flow pumps TBP1010 and TBP1002 are used. Instrument and constant flow pump comes with the protocol; the system includes YSI2700 online analyzer, manual automatic control of constant flow pump, display and storage of on-site data, historical data curve display and fuzzy control strategy functions; (2) adopt the online control system of step (1) to detect the residual sugar concentration, firstly, the computer controls the constant flow pump TBP1010 to extract the fermentation broth, simultaneously controls the constant flow pump TBP1002 to add water for dilution, and then controls the YSI2700 online analyzer to detect the residual sugar concentration; (3) Calculate the lag time of detection according to the extraction rate of the fermentation broth, the length of the pipeline and the analysis time of the YSI2700 online analyzer; (4) The residual sugar concentration in the tank and its rate of change are used as the input of the fuzzy control, and the sugar replenishment rate is used as the output. By determining the input and output language variables of the fuzzy controller, the universe of discourse and the control method, the quantification factor and the scale factor are determined. , the membership function of the input and output of the fuzzy controller, the fuzzy control rules of the fuzzy controller and the clear algorithm steps, the fuzzy control simulation is carried out based on the expert experience, and the fuzzy control parameters are adjusted online to make the residual sugar control effect to be optimal; (5) Apply the designed fuzzy control strategy to the fermentation process: first, analyze the residual sugar concentration through the detection system based on the YSI2700 online analyzer, and then substitute the hysteresis detection value of the residual sugar concentration into the fuzzy controller to obtain the control period the rate of sugar replenishment within the Wherein, the control rules of the fuzzy control strategy are as follows: (a) The output detection value is less than the upper limit value for the first time: assign the initial value g=0, i=1, g is the moment used to judge whether the real-time detection value Cn(i) is greater than the upper limit value a, and i represents the detection moment , output the moment g when the detected value is less than a for the first time; (b) Initial value of the sugar feeding rate: when i<g, Cn(i)>a is always established, the sugar feeding rate is 0; once i>g occurs, the initial value of the sugar feeding rate F(i-1) =(Cn(i-1)-Cn(i))/Δt; (c) Normal sugar supplementation: when the real-time detection value Cn(i)>the upper limit a, the sugar supplementation rate is 0, and the count n starts; when the real-time detection value Cn(i)<lower limit b, judge the The real-time detection value Cn(i-1), on the basis of the sugar supply rate at the last moment or the last 1+n time, add △F as the sugar supply rate, where △F is the fixed sugar supply rate; when the real-time detection value b< When Cn(i)<a, first perform fuzzification processing, and then enter the fuzzy controller, output △u, △u is the sugar supplement rate value that needs to be changed, the internal rules of the fuzzy controller are based on the expert experience and knowledge method, to determine i) The size relationship with 0, and the sugar supplement rate is obtained, thereby obtaining the value E(i) in the fermenter under different sugar supplement rates, E(i)=f[Fn(i)], indicating the fermenter The relationship between the residual sugar concentration and the sugar replenishment rate, and finally determine whether the detection time i is greater than m, and m represents the number of times to be detected.

Images