Solid-state pit whole cycle monitoring management system based on digital twinning technology

Abstract

The application discloses a solid-state tank whole-cycle monitoring management system based on a digital twinborn technology, and relates to the technical field of solid-state fermentation control. The system comprises a data acquisition unit, which acquires and transmits the position and shape size of the tank, the temperature, humidity, ethanol content and carbon dioxide concentration of each tank in solid-state fermentation to a storage unit for classified storage; a data processing unit, which constructs a distribution field model in the tank and imports the established distribution field model into a display unit in real time; and a matching early warning unit, which analyzes and early warns abnormal conditions according to the real-time distribution field state of the distribution field of the tank under the current fermentation time sequence. The various hierarchical structures in the tank are acquired by various sensors, and a visual distribution field model is established, so that the corresponding monitoring indexes in the whole tank are not presented alone, but are distributed and simulated and displayed in the whole space model, so that production personnel can more quickly and intuitively understand the conditions in the tank.

Description

Technical Field

[0001] This invention relates to the field of solid-state fermentation control technology, specifically to a full-cycle monitoring and management system for solid-state fermentation pits based on digital twin technology. Background Technology

[0002] With the continuous expansion of liquor enterprises and the increasing levels of mechanization and automation, intelligent brewing has gradually become an inevitable trend in the reform and development of the liquor industry. However, the diversity of brewing processes, the complexity of the environment, and the density of equipment make comprehensive management of brewing equipment difficult. To increase liquor production, the informatization and intelligent management of automated liquor brewing equipment is an essential part of the transformation and upgrading process.

[0003] Current solid-state fermentation technologies typically employ manual inspections and point sampling to collect relevant data (such as temperature, humidity, and carbon dioxide content) from each fermentation pit. However, the collected data are isolated points with no correlation between them, creating "information silos." This results in insufficient precision in the monitoring and management of solid-state fermentation pits, unreliable data support, and limited guidance value for baijiu production. Therefore, we propose a full-cycle monitoring and management system for solid-state fermentation pits based on digital twin technology. Summary of the Invention

[0004] The purpose of this invention is to provide a full-cycle monitoring and management system for solid-state cellars based on digital twin technology.

[0005] The technical problem solved by this invention is:

[0006] (1) How to collect data on the various levels of the structure in the fermentation pit through multiple sensors and establish a visual distribution field model to solve the problem that the existing solid-state fermentation technology cannot reflect and visualize the data distribution of the various levels of the entire fermentation pit as a whole.

[0007] (2) How to establish a threshold plane in multiple hierarchical structures of the distribution field model in the continuous fermentation time sequence by matching the early warning unit, and the corresponding threshold plane is driven by the threshold mapping function to move up and down in real time, and the normal distribution coefficient is obtained by comparing the volume in the threshold plane with the volume of the corresponding hierarchical structure, so as to solve the problem that the existing technology causes false alarms or the inability to perform overall evaluation and early warning by comparing a single isolated data.

[0008] (3) How to correct the fermentation time sequence by modifying the actual distribution field model of each pit, and substitute the modified fermentation time sequence into the threshold mapping function to correct the threshold plane, so as to solve the problem of increased comparison error caused by inconsistent fermentation time sequence of pits.

[0009] The present invention can be achieved through the following technical solution: a full-cycle monitoring and management system for solid fermentation pits based on digital twin technology, including: a data acquisition unit that collects and transmits in real time the location, shape and size of the pits, as well as the temperature, humidity, ethanol content and carbon dioxide concentration of each pit during solid fermentation, to a storage unit for classified storage;

[0010] The data processing unit constructs a distribution field model of the above data within the cellar and imports each established distribution field model into the display unit in real time for observation and display.

[0011] The matching early warning unit analyzes and warns of abnormal situations in the real-time distribution field status of each fermentation pit under the current fermentation sequence, and notifies the production personnel of the alarm signal for timely remediation.

[0012] The time-series correction unit corrects the time-series deviation between the distribution field of each cellar at the current moment and the standard distribution field simulated under historical data, thereby providing accurate comparison data for the analysis of the matching early warning unit.

[0013] A further technical improvement of the present invention is that the step of the data processing unit constructing a temperature distribution field model within the cellar includes:

[0014] S1: Construct the geometric boundary of the cellar;

[0015] S2: Determine the thermal conductivity, specific heat capacity, and density parameters of the material;

[0016] S3: Divide the entire pit into a hierarchical structure containing multiple flat heat sources;

[0017] S4: Construct the heat conduction equation and boundary condition function based on the corresponding data in steps S1 and S2, thereby forming a set of heat conduction equations within the region of the pit.

[0018] S5: The temperature distribution within the boundary area of ​​the cellar is calculated using the finite element analysis method. Based on the calculation results, different colors are used to mark the temperature distribution field, thus forming a temperature distribution field model.

[0019] A further technical improvement of the present invention is that: the matching early warning unit determines the threshold curves of different types of data in different fermentation sequences within each hierarchical structure based on historical data, fits the threshold mapping function, and establishes two threshold planes in each hierarchical structure of the distribution field model according to different fermentation sequences.

[0020] A further technical improvement of the present invention is that: the matching early warning unit calculates the volume of the distribution field portion located between the upper and lower threshold planes, compares the volume with the overall volume of the corresponding hierarchical structure to obtain the normal distribution coefficient, compares the normal distribution coefficient with the minimum distribution limit, and takes countermeasures based on the comparison results.

[0021] A further technical improvement of the present invention is as follows: when the normal distribution coefficient exceeds the minimum distribution limit, the data of the corresponding distribution field in the hierarchical structure is determined to be normal, and the location data of the data singularity and the corresponding distribution field data are recorded and sent to the storage unit; when the normal distribution coefficient is less than the minimum distribution limit, the data of the corresponding distribution field in the hierarchical structure is determined to be abnormal, a corresponding abnormal alarm is generated, and the generated alarm signal is sent to the equipment terminal of the production personnel through the communication module.

[0022] A further technical improvement of the present invention is that: the timing correction unit uses a relative distribution algorithm to determine the time offset value t between the two distribution fields, and shifts the data of the actual fermentation pit distribution field by t time steps according to the fermentation time sequence. When the variance of the data after the data shift is less than the set value, the timing difference between the two distribution fields is set to t.

[0023] A further technical improvement of the present invention is that the timing correction unit feeds back the obtained timing difference to the matching early warning unit to correct the fermentation timing in the threshold mapping function, thereby constructing a threshold plane corresponding to the corrected fermentation timing in the actual fermentation pit distribution field.

[0024] A further technical improvement of the present invention is that the data processing unit also performs normalization processing on the location data of data singularities occurring in each pit and the corresponding distribution field data recorded in the storage module during the comparison of normal distribution coefficients, and performs correlation analysis together with the production personnel, thereby finding the common conditions for the occurrence of data singularities in the pit from big data processing, and providing guidance for subsequent production.

[0025] Compared with the prior art, the present invention has the following beneficial effects:

[0026] 1. Data is collected from each level of the fermentation pit using multiple sensors, and a visual distribution model is established. This allows the corresponding monitoring indicators within the entire fermentation pit to be distributed and simulated within the entire spatial model, rather than presented in isolation. This enables production personnel to understand the situation within the fermentation pit more quickly and intuitively, and to make timely adjustments.

[0027] 2. By matching the early warning unit, a threshold plane is established in multiple hierarchical structures of the distribution field model within the continuous fermentation time sequence. The corresponding threshold plane is driven by the threshold mapping function to move up and down in real time. The normal distribution coefficient is obtained by comparing the volume in the threshold plane with the volume of the corresponding hierarchical structure. The normal distribution coefficient is compared with the set value, and the corresponding alarm signal is generated based on the comparison result. This avoids the problem of large data processing volume and easy false alarms caused by comparing a single isolated data and the inability to monitor the entire fermentation pit.

[0028] 3. The fermentation timeline of each fermentation pit is corrected by modifying the actual distribution field model of each pit, and the corrected fermentation timeline is substituted into the threshold mapping function to correct the threshold plane, so as to avoid timeline errors in the standard used during comparison and make the comparison results more accurate. Attached Figure Description

[0029] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

[0030] Figure 1 This is a system flowchart of the present invention. Detailed Implementation

[0031] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided.

[0032] Please see Figure 1 As shown, the solid-state cellar full-cycle monitoring and management system based on digital twin technology includes a data acquisition unit, a data processing unit, a matching and early warning unit, a timing correction unit, a storage unit, a communication module, and a display unit.

[0033] The data acquisition unit collects real-time data on the location, shape, size, production personnel, temperature, humidity (i.e., moisture content), ethanol content, and carbon dioxide concentration of each fermentation pit. The collected data is then transmitted to the storage module via wired / wireless transmission through the communication module for classified storage. The data acquisition unit uses timed intervals for data collection and transmission.

[0034] It should be noted that during the data collection process in the fermentation pit, the sampling points in the pit are distributed in a scattered manner, rather than just collecting data in one layer or one area. Therefore, the temperature, humidity, ethanol content and carbon dioxide concentration mentioned above are all related to the corresponding sampling locations in the fermentation pit. The sampling locations are divided into several layers according to the distribution of the main fermentation colonies.

[0035] The data processing unit retrieves data from the storage unit and constructs a distribution field model of the corresponding data within the storage tank. Taking the construction of the temperature distribution field as an example:

[0036] Step 1: Construct the geometric boundary of the pit based on its shape and size information;

[0037] Step 2: Determine the thermal conductivity, specific heat capacity, and density parameters of the material (i.e., mash) inside the fermentation pit based on its properties and quantity.

[0038] Step 3: Since the fermentation colonies at each level are different, the time and intensity of anaerobic respiration are different, and their temperatures are also different. Therefore, the entire pit can be regarded as a hierarchical structure containing multiple flat heat sources. Based on the data in Step 1 and Step 2, the heat conduction equation and boundary condition function are constructed to form a set of heat conduction equations in the pit area.

[0039] Step 4: Calculate the temperature distribution within the boundary area of ​​the pit using the finite element method. Based on the calculation results, color-code the temperature distribution within the entire pit according to the correspondence between high temperature and high color tone.

[0040] Similarly, humidity distribution field models, ethanol content distribution field models, and carbon dioxide concentration distribution field models can be constructed within the fermentation pit;

[0041] Multiple distributed field models generated in real time are sent to the display unit through the communication module, so that production personnel can obtain the current fermentation status of the mash in the fermentation pit;

[0042] The matching and early warning unit obtains the distribution fields of each fermentation pit from the data processing unit, determines the threshold curves of different types of data in different fermentation time sequences within each hierarchical structure based on historical data, and thus fits the threshold mapping function. In the distribution field model of the corresponding data type, the matching and early warning unit establishes a threshold plane in multiple hierarchical structures of the distribution field model according to the fermentation time sequence. This threshold plane is driven by the threshold mapping function and migrates as fermentation time progresses.

[0043] The distribution field state under the current fermentation sequence is analyzed and matched. In a certain hierarchical structure, a triple integral is calculated for the part between the upper and lower threshold planes to obtain the volume between the two threshold planes. The ratio of this volume to the total volume of the hierarchical structure is calculated to obtain the normal distribution coefficient. When the normal distribution coefficient exceeds the minimum distribution limit, the data of the corresponding distribution field in the hierarchical structure is determined to be normal. At the same time, the location data of the data singularity and the corresponding distribution field data are recorded and sent to the storage unit. When the normal distribution coefficient is less than the minimum distribution limit, the data of the corresponding distribution field in the hierarchical structure is determined to be abnormal, and a corresponding abnormal alarm is generated. For example, when the temperature distribution field data is abnormal, a low temperature alarm signal or an over-temperature alarm signal is generated according to the offset direction of the corresponding data. The generated alarm signal is sent to the equipment terminal of the production personnel through the communication module. The equipment terminal is a signal receiving device other than mobile phones.

[0044] The data processing unit also manages the pits under the responsibility of different production personnel by dividing them into zones. Within the area of ​​responsibility of each group of production personnel, each pit is numbered according to its location data and bound to the production personnel of that group. The production personnel information includes the production personnel's years of service and health status. After receiving the alarm signal sent by the matching early warning unit, the corresponding production personnel take remedial measures in a timely manner in response to the alarm signal.

[0045] During the fermentation process of baijiu production, the raw materials of the same batch are basically sealed in the cellar at the same time, which ensures that the fermentation time of the same batch of baijiu is consistent, thus ensuring consistency in the time dimension.

[0046] During the fermentation process of the same batch of baijiu, for the same group of production personnel responsible for the fermentation pits, the distribution field change trend is the same throughout the fermentation cycle under a unified time dimension. The time-series correction unit compares the differences between the actual distribution field state of each fermentation pit and the standard distribution field throughout the entire fermentation cycle:

[0047] Since the distribution field of the above data has time-varying characteristics, it cannot be accurately captured and measured by general mathematical statistical methods. We use the relative distribution algorithm to define the difference between the distribution field of each pit and the standard distribution field of different fermentation time sequences simulated in each level of the historical data.

[0048] In the relative distribution, let S0 be the continuous variable data of the standard distribution field in different fermentation time series, F0(s) be labeled as the standard cumulative distribution function, and f0(s) be labeled as the standard density function; let S be the continuous variable data of the corresponding distribution field of a single fermentation pit in different time series, F(s) be labeled as the actual cumulative distribution function, and f(s) be labeled as the actual density function.

[0049] Let R be the relative probability distribution of S with respect to S0, then the level transformation function with S0 as the baseline is:

[0050] R = F0(S);

[0051] G(r)=F(F0 -1 (r));

[0052] The density function of R is:

[0053]

[0054] Where r represents the percentage of the value, and the value range is [0,1];

[0055] The relative cumulative distribution function G(r) indicates that the proportion of the actual cellar distribution field data domain lagging or exceeding the standard distribution field data domain is lower than the level of the standard distribution field data r=1.

[0056] The relative probability density function g(r) at the r-th quantile of the standard distribution field data domain, F0 -1 (r) When the level is horizontal, the ratio of the frequency of the actual distribution field data of the cellar to the frequency of the standard distribution field data;

[0057] The data of the actual fermentation pit distribution field is mapped to the percentage ranking of the standard distribution field according to the relative cumulative distribution function. For example, in the temperature distribution field, the temperature data of the actual fermentation pit distribution field is converted into a percentage ranking and the position of the temperature data within the entire fermentation time cycle is represented.

[0058] The density ratio between the actual distribution field and the standard distribution field of the cellar is calculated based on the relative probability distribution function, thereby describing the degree of deviation of the actual distribution field of the cellar.

[0059] By combining the relative cumulative distribution function and the relative probability distribution function, the time series offset value t of the two distribution fields is determined. The value of t is negatively correlated with the value of the relative probability distribution function and positively correlated with the value of the relative cumulative distribution function.

[0060] The data domain of the actual fermentation pit distribution field is shifted by t time steps according to the fermentation time sequence, and the variance value of the data after the shift is observed. When the variance value is less than a certain value, the time difference between the two distribution fields is regarded as t. Thus, the fermentation time is corrected during the matching and early warning process, and the corrected time is substituted into the threshold mapping function to establish an accurate threshold plane, which enables more refined control of data monitoring and matching alarms in solid-state fermentation.

[0061] Furthermore, the data processing unit can also perform normalization processing and correlation analysis with production personnel based on the location data of data singularities occurring in each pit in the comparison of normal distribution coefficients recorded in the storage module and the corresponding distribution field data. In this way, it can find the common conditions for the occurrence of data singularities in the pits from big data processing and provide guidance for subsequent production.

[0062] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A full-cycle monitoring and management system for solid-state cellars based on digital twin technology, characterized in that: include: The data acquisition unit collects and transmits data in real time to the storage unit for classified storage, including the location, shape, size of the fermentation pits, and the temperature, humidity, ethanol content, and carbon dioxide concentration of each pit during solid-state fermentation. The data processing unit constructs a distribution field model of the above data within the cellar and imports each established distribution field model into the display unit in real time for observation and display. The steps for constructing a temperature distribution field model within the cellar include: S1: Construct the geometric boundary of the cellar; S2: Determine the thermal conductivity, specific heat capacity, and density parameters of the material; S3: Divide the entire pit into a hierarchical structure containing multiple flat heat sources; S4: Construct the heat conduction equation and boundary condition function based on the corresponding data in steps S1 and S2, thereby forming a set of heat conduction equations within the region of the pit. S5: The temperature distribution within the boundary area of ​​the cellar is calculated using the finite element analysis method. Based on the calculation results, different colors are used to mark the temperature distribution field, thus forming a temperature distribution field model. The matching and early warning unit analyzes and warns of abnormal situations in the real-time distribution field status of each fermentation pit under the current fermentation sequence, and notifies the production personnel of the alarm signal for timely remediation. The matching and early warning unit determines the threshold curves of different types of data in different fermentation sequences within each level structure based on historical data, and fits the threshold mapping function. Based on different fermentation sequences, it establishes two threshold planes in each level structure of the distribution field model. The time-series correction unit corrects the time-series deviation between the distribution field of each pit at the current moment and the standard distribution field simulated under historical data, thereby providing accurate comparison data for the analysis of the matching early warning unit. The timing correction unit uses a relative distribution algorithm to determine the time offset t between two distribution fields. It shifts the data of the actual fermentation pit distribution field by t time steps according to the fermentation time sequence. When the variance of the shifted data is less than the set value, the timing difference between the two distribution fields is set to t. The obtained time difference is then fed back to the matching early warning unit to correct the fermentation time sequence in the threshold mapping function, thereby constructing the threshold plane corresponding to the corrected fermentation time sequence in the actual fermentation pit distribution field.

2. The solid-state cellar full-cycle monitoring and management system based on digital twin technology according to claim 1, characterized in that, The matching early warning unit calculates the volume of the distribution field between the upper and lower threshold planes, compares the volume with the overall volume of the corresponding hierarchical structure to obtain the normal distribution coefficient, compares the normal distribution coefficient with the minimum distribution limit, and takes countermeasures based on the comparison results.

3. The solid-state cellar full-cycle monitoring and management system based on digital twin technology according to claim 2, characterized in that, When the normal distribution coefficient exceeds the minimum distribution limit, the data of the corresponding distribution field in the hierarchical structure is determined to be normal. At the same time, the location data of the data singularity and the corresponding distribution field data are recorded and sent to the storage unit. When the normal distribution coefficient is less than the minimum distribution limit, the data of the corresponding distribution field in the hierarchical structure is determined to be abnormal. A corresponding abnormal alarm is generated, and the generated alarm signal is sent to the equipment terminal of the production personnel through the communication module.

4. The solid-state cellar full-cycle monitoring and management system based on digital twin technology according to claim 1, characterized in that, The data processing unit also performs normalization processing on the location data of data singularities occurring in each pit and the corresponding distribution field data recorded in the storage module during the comparison of normal distribution coefficients. It then performs correlation analysis together with the production personnel to find the common conditions for the occurrence of data singularities in the pits from big data processing, providing guidance for subsequent production.

Images