Cooking and drying integrated control method for parboiled rice

Abstract

The invention discloses a preboiled rice cooking and drying integrated control method, particularly relates to the technical field of grain processing process control, and is used for solving the problems that the product quality fluctuation is large and stability and energy consumption optimization are difficult to consider due to the fact that existing preboiled rice cooking and drying procedures are dispersed and dependent on experience parameter adjustment. Initial moisture content, variety, granularity and soaking parameter eigenvectors are introduced in a batch level, a heat and humidity process model is established in combination with process parameters of a cooking section and each drying section, heat and humidity process fingerprints are generated under unified time service, and a cooking saturation index and a drying stress load index are calculated; and process parameters are adjusted in a linkage mode according to the stress quota and the energy consumption quota, so that batch-level dynamic closed-loop control aiming at different unprocessed grains and working conditions is achieved within the process capacity range of the production line, the water content, hardness and color of the preboiled rice finished product are stable and close to the target, and the risks of dry cracks and insufficient gelatinization are reduced.

Description

Technical Field

[0001] This invention relates to the field of grain processing control technology, specifically to an integrated control method for steaming and drying parboiled rice. Background Technology

[0002] In existing parboiled rice production lines, the cooking and multi-stage drying processes are mostly controlled by individual equipment or based on experience-based formulas. Typically, a fixed combination of cooking temperature, holding time, and drying temperatures for each stage is predetermined during the process design phase for a specific rice variety and initial moisture content. During actual operation, adjustments are made only based on operator experience or minor online variations in moisture and temperature. Differences in the initial moisture content, particle size distribution, and soaking conditions of batches of raw grain are often only recorded discretely in incoming material inspection or ledgers, without forming a systematic link with subsequent cooking and drying control strategies. Process data such as temperature, humidity, and moisture content in the cooking and drying stages are also stored piecemeal by equipment and by time, lacking a unified "full-process trajectory" description linked by batch and time axis, and even more so, lacking a unified index system for quantitatively evaluating the degree of gelatinization and drying stress level.

[0003] Under the aforementioned control and recording model, parboiled rice production often struggles to grasp in a timely and accurate manner "the degree to which a particular batch of raw grain has been cooked under current operating conditions, whether the drying stress is approaching the crack risk boundary, and how the remaining processes can be allocated and balanced between quality and energy consumption." This leads to common engineering phenomena: to prevent insufficient gelatinization and excessive cracking, there is a general tendency to conservatively increase the cooking and drying intensity. While the product quality can basically meet the standards, batch-to-batch fluctuations in hardness and color still exist, and energy consumption levels are difficult to continuously optimize. Once the moisture content of the raw grain, seasonal temperature and humidity, or equipment conditions deviate, the original experience-based formula often requires a long period of repeated adjustments to regain stability. At the same time, existing technologies for recording the operation process and final quality results of each batch are mostly limited to simple ledgers and a few statistical indicators. There is a lack of a mechanism to link batch characteristics, process thermal and humidity history, control parameters, and quality results in a long-term manner through versioning and evidence chains for subsequent process correction and strategy updates. This makes it difficult to systematically distill experience across seasons and raw grain sources into reusable control knowledge.

[0004] Therefore, under the existing integrated production of parboiled rice cooking and drying, the core technical problem can be summarized as follows: the lack of a control and management method based on batch characteristics that can characterize the cooking and multi-stage drying heat and moisture process on the time axis and link the operation process with the quality results in a closed loop. This makes it difficult for the production line to adjust the process parameters of each stage in a timely and quantitative manner and to gradually revise the process model and control strategy when the raw grain and operating conditions change. It is also difficult to simultaneously take into account the cross-batch stability of parboiled rice product quality and the long-term optimization of energy consumption level. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an integrated control method for parboiled rice cooking and drying, thereby solving the problems mentioned in the background section.

[0006] To achieve the above objectives, the present invention provides the following technical solution: an integrated control method for parboiled rice cooking and drying, comprising:

[0007] S1. Collect the initial moisture content, variety, particle size and soaking parameters of raw grains in batches, generate batch feature vectors and bind them to the target finished product moisture content, brown rice hardness and color index;

[0008] S2. Based on the batch feature vector and the process parameters of the cooking section and each drying section, establish an integrated heat and humidity process model for cooking and drying, generate a heat and humidity process template that meets the target indicators offline, and solidify the model parameters.

[0009] S3. Divide the operation time slices under unified time synchronization, collect the temperature, humidity, moisture and residence time of the cooking section and each drying section according to the time slice, and use the thermal and humidity history model to obtain the thermal and humidity history fingerprint of this batch.

[0010] S4. Calculate the cooking saturation index and drying stress load index based on the thermal and wet history fingerprint, combine them with the target index to form the mass offset, and determine the stress quota and energy consumption quota for the remaining cooking and drying processes.

[0011] S5. Based on the stress quota and energy consumption quota, the steam temperature, steam pressure, holding time of the cooking section, as well as the air inlet temperature, air volume, and conveying speed of each drying section are adjusted in a coordinated manner so that the heat and humidity process fingerprint follows the heat and humidity process template.

[0012] S6. Record the feature vectors, thermal and humidity history fingerprints, control parameters and final quality results of each batch, periodically correct the thermal and humidity history model and thermal and humidity history template parameters, and form an integrated control strategy for cross-batch self-learning.

[0013] Furthermore, S1 includes:

[0014] A batch identifier consisting of date, time, warehouse number, and supplier code is generated for each batch of raw grain entering the production line;

[0015] The initial moisture content, variety code, particle size grade, soaking time and soaking water temperature of the batch of raw grains are collected and used to construct a batch feature vector according to preset fields and coding rules.

[0016] According to the variety process specifications, the target finished product moisture content, brown rice hardness and color index are bound to this batch, and the process calculation unit writes the batch feature vector and target index along with the process rule version number into the batch database.

[0017] Furthermore, S2 includes:

[0018] The process calculation unit establishes an integrated heat and humidity process model for cooking and drying based on batch feature vectors, cooking section process parameters, and each drying section process parameters.

[0019] The grain layer thickness was divided into multiple representative locations, the running time was divided into fixed-length time slices, the model parameters were adjusted and the model version number was set according to the working condition records.

[0020] Based on the target finished product's moisture content, brown rice hardness, and color indicators, a heat and humidity history template is generated. The correspondence between the template identifier and the variety code and model version number is established, and the template identifier is stored in the template library.

[0021] Furthermore, S3 includes:

[0022] Under a unified time synchronization standard, continuous production time is divided into run time slices with run time slice numbers;

[0023] The on-site control equipment collects the temperature, humidity, moisture content and residence time of the cooking section and the drying section during each operating time segment, and generates operating condition transmission messages carrying batch identifiers and section numbers.

[0024] The process calculation unit aligns and cleans the operating condition reports according to the running time slice number, calls the integrated cooking and drying heat and moisture history model to estimate the temperature and moisture content of representative locations, and organizes the estimation results into heat and moisture history fingerprints and writes them into the running library.

[0025] Furthermore, S4 includes:

[0026] Based on the thermal and moisture history fingerprint in the runtime library, the process calculation unit identifies the set of running time slices corresponding to the cooking section and each drying section. Within each running time slice, it calculates the cooking saturation index and drying stress load index based on the temperature and moisture content of representative locations. The cooking saturation index and drying stress load index are then compared with the target finished product moisture content, brown rice hardness, and color index of the batch to generate a quality offset.

[0027] Furthermore, the process calculation unit performs statistical analysis on the mass offset within the observation window consisting of continuous operating time slices and identifies the trend of mass offset changes. Based on the degree of deviation of the cooking saturation index and drying stress load index from the target range, combined with the equipment safety process window and energy consumption budget, it determines the stress quota and energy consumption quota for the remaining cooking section and each drying section of this batch.

[0028] Furthermore, S5 includes:

[0029] The process calculation unit registers the equipment safety process window and allowable variation range for the process parameters of the cooking section and drying section in the equipment parameter library;

[0030] When the mass offset reaches the adjustment threshold, the parameter change is calculated under the constraints of stress quota and energy consumption quota, and an instruction message containing instruction number, batch identifier, parameters and strategy version number is generated and sent to the field control equipment through the industrial network.

[0031] The field control equipment identifies duplicate messages and sequence number abnormalities based on the recorded execution instruction sequence number. When the parameters exceed the equipment's safe process window, it performs boundary checks and generates error code feedback information.

[0032] Furthermore, S6 includes:

[0033] After each batch is completed, the process calculation unit writes the batch feature vector, thermal and humidity history fingerprint, key operating time slice process parameters, control instruction execution records and final quality results of that batch into the runtime library and historical batch library in a field-based record format.

[0034] The batch identifier, thermal and humidity history model version number, thermal and humidity history template version number, process rule version number, and control strategy version number are written into the runtime library and historical batch library in a field-based manner to form the evidence chain record for that batch.

[0035] Furthermore, the process calculation unit groups the batch records according to variety and raw grain conditions within a preset time period, performs statistical analysis on the thermal and moisture history fingerprints of each group and the final quality results, and determines the model parameters and thermal and moisture history template parameters corresponding to the stable operating conditions.

[0036] The updated model parameters and thermal-humidity history template parameters are registered in the model library and template library as new versions. At the same time, a version locking strategy is adopted to make the new version effective in the batch launched after the version switch, so as to achieve cross-batch self-learning optimization.

[0037] Compared with the prior art, the present invention has the following beneficial effects:

[0038] 1. By introducing batch feature vectors of initial moisture content, variety, particle size, and soaking parameters at the batch level, and establishing an integrated heat and moisture history model for cooking and drying in conjunction with the process parameters of the cooking section and each drying section, a heat and moisture history fingerprint is continuously generated based on a unified time slice, and the cooking saturation index and drying stress load index are calculated. Under the constraints of stress quota and energy consumption quota, the cooking and drying process parameters are adjusted in a coordinated manner, thereby achieving batch-level dynamic closed-loop control for different raw grains and working conditions within the actual production line process capacity. This ensures that the moisture content, hardness, and color of the parboiled rice product stably follow the target indicators, reduces fluctuations caused by human experience intervention, and lowers the risk of drying cracks and insufficient gelatinization.

[0039] 2. By structurally recording the batch feature vectors, thermal and moisture history fingerprints, key process parameters, control command execution status, and corresponding model, template, and process rule version numbers of each batch throughout the entire production process, a traceable evidence chain is formed. Statistical analysis of the results of multiple batches is conducted according to a preset evaluation cycle. Based on the grade rice yield, energy consumption level, and proportion of abnormal batches, the thermal and moisture history model parameters, thermal and moisture history template trajectory bands, and stress quota rules are version-modified and locked for application. This achieves continuous optimization of the integrated cooking and drying control strategy under conditions of cross-seasonal, cross-grain source, and equipment status changes, enabling the production line to have self-learning and self-adaptive capabilities, while taking into account quality stability, energy consumption level, and process compliance and traceability. Attached Figure Description

[0040] Figure 1 This is a schematic diagram of the process of an integrated control method for steaming and drying parboiled rice according to the present invention. Detailed Implementation

[0041] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0042] Example: Figure 1 A flowchart illustrating an integrated control method for parboiled rice cooking and drying according to the present invention is provided. The integrated control method for parboiled rice cooking and drying includes:

[0043] S1. Collect initial moisture content, variety, grain size, and soaking parameters of raw grains in batches, generate batch feature vectors, and bind them to the target finished product moisture content, brown rice hardness, and color indicators. The specific implementation is as follows:

[0044] On a parboiled rice cooking and drying integrated production line, in order to enable the subsequent integrated cooking and drying control to distinguish different incoming materials and make targeted process adjustments for each batch, a batch identifier is established for each batch of raw grain entering the line. The batch identifier is used to uniquely identify a batch of raw grain throughout the entire production process. It can be formed by combining the date, time, warehouse number, and supplier code in an agreed order. Through this identifier, the same number is always referenced in each subsequent process and link to avoid confusion when recording and calling the same batch. After the batch identifier is established, the initial moisture content, variety, particle size, and soaking parameters of the batch of raw grain are collected. These quantitative information reflecting the physical properties and pretreatment status of the raw grain are organized into a batch feature vector. The batch feature vector is an ordered set composed of several fields. Each field corresponds to a physical quantity and uses a clear unit and code, which is used for unified calling in subsequent modeling and control calculations.

[0045] The initial moisture content is the proportion of water contained in the raw grain by mass before entering the cooking process. In this field, online or offline moisture measuring devices are typically used for measurement. To balance real-time performance and representativeness in continuous production, an online moisture measuring device installed on the feed hopper or conveyor belt can be used to obtain the initial moisture content. This device collects electromagnetic or infrared characteristics as the raw grain passes through the measurement area and obtains the moisture value according to a pre-calibrated conversion relationship. The moisture value is uniformly expressed as a percentage by mass, and a timestamp in seconds is appended to the data collection, recorded together with the batch identifier to ensure time consistency. The variety field is used to indicate the rice variety to which this batch of raw grain belongs. To avoid ambiguity caused by using natural language descriptions in calculations and queries, the variety uses a variety code uniformly maintained internally by the enterprise. The table is coded, and this coding table is pre-configured in the system and associated with supplier information and variety characteristics; the particle size field is used to describe the particle size distribution of the batch of raw grain. In this field, particle size grades are usually given by screening tests. In continuous production scenarios, online screening devices can be configured or periodic sampling can be performed. The screening results are mapped to the predefined particle size grade code, and the value range of the particle size grade is consistent with the commonly used specifications of the enterprise; the soaking parameters are used to describe the soaking treatment of the batch of raw grain before entering the cooking process, including at least soaking time and soaking water temperature. The soaking time is recorded by a timing device set on the soaking tank, and the soaking water temperature is recorded by a temperature sensor installed in the soaking tank. These two quantities use a unified clock as the time base, and the values ​​are solidified into the batch feature vector at the beginning and end of the batch.

[0046] To ensure clear control objectives and facilitate comparison with actual quality inspection results, target finished product moisture content, brown rice hardness, and color indicators are assigned to each batch based on the company's variety process specifications and specific order requirements, while establishing the batch feature vector. The target finished product moisture content is the range that the batch of parboiled rice should reach when leaving the factory. The brown rice hardness is the breaking force range measured by a texture analyzer or other mechanical testing equipment on the milled brown rice. The color indicator is the grade range determined by the yellowness value measured by a colorimeter or the company's color grade standard. These target indicators are pre-specified in the process documents and correspond to the variety code and product grade, and are registered in the system in the form of rules.

[0047] In actual system implementation, all initial moisture content, variety code, grain size grade, soaking parameters, and target finished product moisture content, brown rice hardness, and color indicators related to the batch are written into the batch database by the process calculation unit in a field-separated form. The batch database is a storage unit that centrally stores the characteristics and target indicators of each batch. It can be set up on the industrial computer or process server where the process calculation unit is located. The process calculation unit can be implemented using an industrial computer with data storage and computing capabilities. In addition to the above fields, each batch record also includes a process rule version number, record time, and operator identifier to indicate the process rule version used in the batch and form a traceable chain of evidence.

[0048] During the transmission of batch feature vectors from the acquisition layer to the process calculation unit, the control equipment of the acquisition layer summarizes the measured values ​​through the fieldbus and then sends them to the process calculation unit through the industrial Ethernet. The fields in the message structure are organized according to the pre-agreed order and data type. The message carries the batch identifier and interface version information. After receiving the message, the process calculation unit checks the integrity of the fields and the range of values, and returns confirmation information to the acquisition layer.

[0049] If no confirmation is received within the preset number of attempts, the acquisition layer records the transmission process as a failure and writes an error status flag in the corresponding batch record, such as recording error code E04 and the number of failures. The preset number of attempts can be set to three or an integer value close to three. The specific value is uniformly configured by the enterprise during system initialization according to network stability and real-time requirements, so that those skilled in the art can maintain consistent behavior.

[0050] When the process calculation unit reads the batch feature vector in subsequent steps, it can identify this state. In the absence of reliable information on certain fields, it selects a conservative combination of target indicators for the batch according to a pre-set safety mode. Preferably, the target combinations that have been verified in the process specification for the same variety can be sorted according to quality grade and energy consumption requirements. When the information is incomplete, a target combination with a lower quality requirement or a larger energy consumption reserve can be selected. Alternatively, a set of target finished product moisture content, brown rice hardness, and color indicators with a high pass rate in historical statistics can be selected as the conservative target indicators for the batch to reduce the risk to subsequent quality stability caused by the lack of information in the early stage.

[0051] Through the above arrangements, a clear implementation path is provided for various aspects, including the composition of batch identifiers, the physical quantities and their measurement methods and units included in the batch feature vector, the storage method of batch features and target indicators in the batch database, and the communication mechanism between the acquisition layer and the process calculation unit. This enables those skilled in the art to establish batch feature vectors for each batch of raw grain and bind target finished product moisture content, brown rice hardness, and color indicators within the actual production line's process capacity and applicable range, based on the online measurement and industrial communication infrastructure of the parboiled rice production line. This provides stable basic data for subsequent integrated modeling and control of steaming, cooking, and drying based on batch features.

[0052] For example, under a representative production condition, the rated processing capacity of the production line can be set to 20 tons per hour, the initial moisture content can be set to the range of 14% to 16%, the soaking time can be set to 4 to 6 hours, the soaking water temperature can be set to 45 to 55 degrees Celsius, the target finished product moisture content can be set to 13% to 13.5%, and the hardness and color index of brown rice can be set to a certain combination of grades in the enterprise standard. Under this condition, the process calculation unit completes the establishment of batch feature vectors and binding of target indicators in the above manner. After being permanently stored in the batch library, it can be called by subsequent steps, so that engineers can reproduce the implementation process of this step without creative labor.

[0053] S2. Based on batch feature vectors and process parameters of the cooking and drying sections, establish an integrated cooking and drying thermo-humidity history model, generate offline thermo-humidity history templates that meet target indicators, and solidify model parameters. The specific implementation is as follows:

[0054] To implement unified and adjustable integrated cooking and drying control for different batches within the actual parboiled rice production line, the process calculation unit establishes an integrated cooking and drying thermal and moisture history model based on batch feature vectors and process parameters of the cooking section and each drying section.

[0055] The integrated heat and moisture process model for cooking and drying is used to describe the relationship between the internal temperature and moisture content of grains and time under given process conditions. The cooking section and multiple drying sections are regarded as a continuous heat and moisture treatment path. By simplifying the heat transfer and moisture migration along this path, those skilled in the art can predict the temperature and moisture content of grains over time using a limited number of process parameters without directly solving complex physical equations.

[0056] Therefore, in the process calculation unit, the cooking steam temperature, steam pressure, holding time, filling thickness, and material quantity per unit time related to the cooking section are organized into cooking section process parameters. The inlet air temperature, outlet air temperature, outlet air humidity, air volume, material layer thickness, and conveying speed of each drying section are organized into drying section process parameters. The cooking section process parameters reflect the ability of the cooking equipment to transfer heat and moisture to the grain, while the drying section process parameters reflect the influence of each drying section on moisture evaporation and temperature changes.

[0057] When establishing the model, it is preferable to divide the grain layer into multiple representative positions in the thickness direction. The number of representative positions can be set to three to ten, which are used to represent the surface, near-surface, and internal regions of the grain layer, respectively. It is assumed that the temperature and moisture content at each representative position can be approximately uniformly distributed within a time slice. Then, the model is divided into continuous small time slices, and the length of the time slice can be set to one minute or the same as the running time slice. Within each time slice, the change in grain temperature and moisture content within the time slice is calculated based on the temperature and moisture content of the previous time slice and the corresponding cooking or drying process parameters. This method of segmented calculation in space and time is a commonly used engineering simplification method in this field, which is beneficial to control the amount of calculation while ensuring that the calculation accuracy meets the process requirements.

[0058] To ensure that the model parameters conform to the characteristics of the actual production line, during the commissioning and stable operation phases of the production line, process engineers select no fewer than three representative batches. For each representative batch, the working conditions and final quality results are continuously recorded from the beginning to the end of the batch. The working conditions record consists of operational information arranged in chronological order, including the batch feature vector established according to step S1, the time-related process parameters of the cooking section and each drying section, and the material temperature, moisture content, and exhaust humidity measured at key locations. The final quality results include the moisture content, brown rice hardness, and color grade of the corresponding batch of finished product.

[0059] In the process calculation unit or dedicated analysis environment, process engineers compare the prediction results of the integrated cooking and drying heat and moisture history model on these representative batches with the actual material temperature, moisture content and final quality results in the operating condition records. By adjusting the parameters in the model used to characterize the heat transfer intensity, moisture migration rate and equipment characteristics, the prediction curve is made close to the actual measurement results within a reasonable error range. The error range can be set to a temperature deviation within one to three degrees Celsius and a moisture deviation within 0.5% to 1.5% of the mass fraction. The specific values ​​are uniformly agreed upon by the enterprise in the process specification according to the process control accuracy requirements.

[0060] Through the above method, a set of model parameters applicable to the current production line equipment layout and product characteristics are obtained, and then solidified in the process calculation unit in the form of an integrated cooking and drying heat and humidity process model. When the model is solidified, a model version number is assigned. The model version number is used to distinguish the model formed in different adjustment stages and is stored in the model library along with the model.

[0061] To apply the model to the control of specific quality targets, after the model parameters stabilize, process engineers call the model in an offline environment and combine it with different varieties, different target finished product moisture content, brown rice hardness and color indicators to calculate the target trajectory of grain internal temperature and moisture content along the time axis in the cooking section and each drying section, and organize these target trajectories into a thermal and moisture history template.

[0062] The heat and moisture history template is a trajectory band used for reference in subsequent control steps. It is used to characterize the temperature and moisture content changes that grains should undergo when reaching a certain target quality combination. Its content preferably includes target temperature ranges and target moisture content ranges at multiple time points throughout the entire process from entering the cooking section to leaving the final drying section.

[0063] For ease of use in engineering, when generating the thermal and moisture history template, the time axis can be divided into multiple time periods according to the process rhythm, such as by minute or by running time slice. Within each time period, the target range of internal temperature and moisture content of the grain can be given, instead of giving precise values ​​at the second level. This segmented trajectory is easy to store and recall while ensuring control effect.

[0064] For example, in a preferred embodiment, for a parboiled rice variety with medium grain size and appropriate initial moisture content, the cooking stage in the heat and moisture process template can be set to maintain the internal temperature of the grains within the range of 90°C to 100°C for 20 to 30 minutes to ensure sufficient starch gelatinization. Subsequently, in the first drying stage, the air inlet temperature is set to 110°C to 120°C to rapidly dehydrate the surface of the grains while the internal moisture content decreases slowly. In the second drying stage, the air inlet temperature is set to 90°C to 100°C to further reduce moisture and decrease the moisture gradient. In the third drying stage, the air inlet temperature is set to 75°C to 85°C for gentle drying to approach the target finished product moisture content. The heat and moisture process template obtained in this way is a target trajectory band under the variety and the target quality combination.

[0065] For each heat and humidity process template, the system registers the template identifier, associated variety code, applicable target finished product moisture content range, brown rice hardness range and color grade combination in the template library, as well as the corresponding integrated cooking and drying heat and humidity process model version number, and records the representative trial batch number and laboratory test report number on which the template was based, as part of the chain of evidence.

[0066] Template and model libraries can be stored in the same industrial computer or process server as the process calculation unit. Different varieties, different target combinations, and different versions can be distinguished by hierarchical directories or tags. Any modification to model parameters or thermal and humidity history templates is done by adding a new version without overwriting the original version. This ensures that the model and template used in a certain period of time can be traced back based on the version number.

[0067] To select a suitable heat and humidity history template in actual production, when the batch feature vector reaches the process calculation unit, the process calculation unit searches the template library for heat and humidity history templates that meet the combination conditions based on the variety code and the target finished product moisture content, brown rice hardness, and color index. Preferably, the templates that cover the target index range in the template set that completely matches the variety code are first screened, and then one that has been verified as a stable version is selected as the master control template for the current batch according to the version number. If multiple templates meet the conditions, one can be selected as the master control template for the current batch according to the priority rules preset in the process specification.

[0068] In cases where a perfectly matching template cannot be found, such as when a new variety is introduced and a dedicated template has not yet been formed, a conservative template of the same type of variety can be preferred. The determination of the same type of variety is based on the variety grouping information in the variety coding table, as well as conditions such as particle size grade and initial moisture content range. Before system deployment, the process engineer shall pre-define this in the form of a static mapping table in the process specification and manage it with a version number, so that the same variety code always corresponds to a unique set of templates of the same type of variety during operation. When the process calculation unit selects a template of the same type of variety, it records this selection process and the identifier of the replaced template as a template substitution relationship and stores it in the evidence chain so that the template can be updated after supplementing the trial data for this type of batch in the future.

[0069] The selected thermal and humidity history template and its version number are stored together with the current batch identifier in the batch library or runtime library in the batch additional information field, which serves as the target reference for calculating thermal and humidity history fingerprints and adjusting process parameters in subsequent steps.

[0070] Based on this, those skilled in the art can, within the actual production line's process capacity and applicable range, configure measurement and control methods for the cooking and drying sections, and realize the establishment, parameter tuning, and generation and management of the integrated cooking and drying thermal and moisture history model on the process calculation unit. In actual operation, the appropriate template can be automatically matched according to the batch feature vector, thereby providing a reproducible and traceable basis for subsequent dynamic control.

[0071] S3. Under unified time synchronization, divide the operation into time slots, and collect temperature, humidity, moisture content, and residence time of the cooking section and each drying section according to the time slot. Use the thermal and humidity history model to obtain the thermal and humidity history fingerprint of this batch. The specific implementation is as follows:

[0072] In order to break down the continuous production process into time units that are easier to analyze and control in the actual integrated parboiler cooking and drying production line, the production line establishes a unified time reference between the process calculation unit and the field control equipment through a clock synchronization mechanism. The unified time reference means that the process calculation unit provides a unified time source and distributes the current time to each field control device through the industrial network. Each device corrects its local clock according to the received time information, so that the time representation on different devices at the same moment is consistent, thereby ensuring that the data from different measuring points can be aligned on the same time axis.

[0073] Based on unified time synchronization, continuous production time is divided into multiple running time slices. Each running time slice is a time interval of equal length under a unified time reference. Each time interval corresponds to a number. The length of the running time slice can be set to sixty seconds or the same as the length of the time slice used in the integrated cooking and drying thermal and wet process model in step S2, so as to form a one-to-one correspondence in the time dimension. The number of each running time slice and the second-level timestamp of the start time of the time slice together constitute a time positioning identifier, which is used to uniquely mark the time slice in subsequent recording and calculation.

[0074] During each operating time slice, the field control equipment of the cooking section and each drying section collects the temperature, humidity, moisture and residence time related to this section according to a predetermined rhythm. The field control equipment refers to the control unit that is directly connected to the field sensors and actuators and has the functions of signal acquisition and parameter adjustment, such as a programmable controller or embedded control device.

[0075] Temperature acquisition points in the cooking section can be arranged at representative locations within the steam medium channel and the grain layer. The former reflects the temperature level of the cooking medium, while the latter reflects the internal heating of the grain. The humidity in the cooking section is provided by humidity monitoring devices located in the steam pipeline and equipment casing, expressed as a percentage of relative humidity. The moisture content in the cooking section is estimated by an online moisture measurement device at the cooking outlet, expressed as a percentage of mass fraction. The residence time in the cooking section is calculated by the on-site control equipment based on the material's travel length and flow rate within the cooking equipment. The material flow rate can be converted based on the conveyor speed and equipment structural parameters. The residence time is expressed in seconds.

[0076] Temperature sampling points in each drying section include two types: air inlet and air outlet. The air inlet temperature reflects the hot air supply conditions, while the air outlet temperature and humidity together reflect the heat and humidity exchange during the drying process. Humidity sampling points in the drying section are arranged in the exhaust channel and expressed as a percentage of relative humidity. The moisture content in the drying section is periodically measured by an online moisture measuring device at the outlet of that section. The moisture measurement cycle can be set to two to five operating time segments. At least one online moisture measurement is completed within each moisture measurement cycle, and the measured moisture value is correlated with the most recent operating time segment. The residence time in the drying section is calculated by converting the speed of the conveyor device and the material travel length of that section, and is expressed in seconds.

[0077] At the end of each operating time slot, the above quantities are aggregated in the field control equipment. Aggregation involves statistically organizing multiple raw measurements that may be collected from the same measurement point within that time slot. For example, the average or median of temperature and humidity sampling values ​​is calculated, and when a moisture measurement occurs once in this time slot, the record is directly used. These aggregated values ​​are then combined with the time slot number, timestamp, equipment number, batch identifier, and segment number to form a working condition transmission message.

[0078] Operating condition transmission messages are operation information messages sent from field control equipment to the process calculation unit. The order and data type of the fields in the message are pre-defined in the interface specification and identified by the interface version number to maintain compatibility when the interface is adjusted.

[0079] Under normal circumstances, the field control equipment generates a status report message after each operating time slice and sends it to the process calculation unit from bottom to top via fieldbus and industrial Ethernet. When a critical measuring device fails to provide a valid reading within a specified time window during an operating time slice, such as a temperature sensor returning a fixed value beyond its range or a moisture measuring device failing to return a valid result in multiple consecutive samplings, the field control equipment identifies this status as indicating that the measuring point has no valid data within that time slice. It then attaches error code E01 to the status report message corresponding to that time slice, along with the corresponding batch identifier and device number. Error code E01 is used to indicate the presence of measurement gaps during subsequent processing.

[0080] After receiving the operating condition transmission message, the process calculation unit first aligns the data from different equipment and different segments according to the time slice based on the timestamp and running time slice number in the message. Data from the same batch and the same running time slice are aggregated into a complete set of operating condition information. After alignment, the process calculation unit performs a reasonableness check on each measurement value, discarding values ​​that clearly violate physical boundaries. For example, temperatures falling below zero degrees Celsius or above 150 degrees Celsius, relative humidity falling below zero or above 100, and moisture values ​​falling within extremely low or high ranges that are impossible for grain to have are considered invalid and not included in subsequent calculations. For cases with a small number of missing measurements within a limited time slice, such as continuous... If a moisture value is missing for a certain period within one to three operating time slices, the process calculation unit can interpolate based on the changing trend of that measurement point in the preceding and following time slices. For example, it can use the average value of adjacent time slices or a linear changing trend to estimate the representative value of that time slice and internally record that time slice as the interpolated time slice for subsequent analysis to identify the reliability of the information. When the number of consecutive missing time slices exceeds a preset upper limit, such as more than three consecutive operating time slices, the process calculation unit can mark that period as having significantly decreased data reliability and no longer perform fine-tuning on that period. Instead, it can be handled by the safety mode. The safety mode is used to keep the process parameters within a conservative operating range when the data reliability has significantly decreased, prioritizing the maintenance of production continuity and avoiding the implementation of new quota adjustment strategies.

[0081] After data alignment and cleaning are completed, the process calculation unit calls the integrated cooking and drying thermal and moisture history model locked in step S2. Based on the process parameters of each segment in the current running time slice and their corresponding measured temperature, humidity and moisture content, it estimates the temperature and moisture content distribution of each representative position inside the grain in the current batch within the running time slice. The estimation uses the same representative position division and time slice length as when the model was established. Starting from the temperature and moisture content state of the previous time slice, the temperature and moisture content changes caused by the process conditions of the cooking section or each drying section are superimposed in the current time slice to obtain the internal state of the grain at the end of the current time slice, and compares it with the thermal and moisture history template matched for the batch in step S2.

[0082] The above estimation results are organized to form the thermal and humidity history fingerprint of the batch in the current operating time slice. The thermal and humidity history fingerprint is a summary information used to characterize the actual thermal and humidity history of the batch on the time axis. It preferably includes the temperature value and moisture content value at each representative location, as well as the degree of deviation of these values ​​from the target range given by the corresponding thermal and humidity history template. The degree of deviation can be described by the direction and magnitude of deviation, and is used to determine whether the batch deviates from the predetermined trajectory in subsequent steps.

[0083] The generated thermal and humidity history fingerprint, along with the batch identifier, runtime slice number, model version number, and template version number, is written into the runtime library. The runtime library is a storage unit used to store the operating condition information and calculation results of each batch in each runtime slice. It is usually set up on the industrial computer or process server where the process calculation unit is located, and is stored in a field-based record method.

[0084] By dividing the operating time slices according to a unified time reference, collecting and aggregating the temperature, humidity, moisture content and residence time of the cooking and drying sections within each time slice, and sending the data to the process calculation unit via operating condition reports and recording the thermal and moisture history fingerprint in the operating database, those skilled in the art can continuously reconstruct and quantify the thermal and moisture history of each batch on the time axis within the actual process capabilities and applicable scope of the parboiled rice production line. This provides a reproducible and traceable basis for subsequent calculation of the cooking saturation index, drying stress load index and execution of linkage adjustments.

[0085] For example, in a preferred embodiment, the running time slice length can be set to sixty seconds, and the time synchronization cycle of unified time can be set to one to five minutes. On a production line with a rated processing capacity of twenty tons per hour, the cooking section can be equipped with two to three temperature acquisition points and one online moisture measurement device. Each drying section can be equipped with one temperature acquisition point for the air inlet and one humidity acquisition point for the air outlet, as well as a moisture measurement device for the discharge port. After each time slice during operation, the field control equipment generates a working condition transmission message in the above manner and forms a continuous thermal and moisture history fingerprint sequence in the process calculation unit. Based on this, engineers can reproduce the calculation process and control basis of this step.

[0086] S4. Calculate the cooking saturation index and drying stress load index based on the thermal and moisture history fingerprint, combine them with the target indicators to form a quality offset, and determine the stress quota and energy consumption quota for the remaining cooking and drying processes. The specific implementation is as follows:

[0087] The process calculation unit quantifies the state of each batch during cooking and drying based on continuously recorded thermal and moisture history fingerprints in the runtime database. It calculates the cooking saturation index and drying stress load index, and compares these with pre-defined target finished product moisture content, brown rice hardness, and color indicators for that batch, generating a quality offset. The unit then statistically analyzes the trend of this quality offset within a certain observation window. Combining this with the current cooking progress and remaining drying time, it calculates the stress and energy consumption quotas for that batch in subsequent cooking and drying processes. The cooking saturation index, a dimensionless indicator, describes the degree of gelatinization of the grains in the current batch during the cooking stage. It is calculated by analyzing the temperature changes over time at representative locations during the cooking stage in the thermal and moisture history fingerprint.

[0088] The process calculation unit identifies the set of operating time slices corresponding to the cooking stage in the thermal and wet process fingerprint. Within each time slice, it detects whether the internal temperature of the grain at each representative position is within the gelatinization temperature range corresponding to the variety. The gelatinization temperature range is given in advance according to the variety in the process specification. For example, the gelatinization temperature range of a certain variety can be set to 80 degrees Celsius to 95 degrees Celsius. For time slices that are within the gelatinization temperature range, the cumulative duration of the internal temperature of the grain within the range is counted, and the temperature difference between the inside and the surface is calculated. When the temperature stays in the gelatinization temperature range for a long time and the temperature difference between the inside and the surface does not exceed the pre-set upper limit of the temperature difference (e.g., five degrees Celsius), the degree of gelatinization is considered to be high.

[0089] The process calculation unit maps the cooking time to the total cooking time, the degree to which the temperature approaches the upper limit of the gelatinization range, and whether the temperature difference meets the constraints, into a cooking saturation index between zero and one. The closer the index value is to one, the more complete the gelatinization. Preferably, the process specification can stipulate that the cooking saturation index of a certain main variety should be controlled within the range of 0.7 to one, in order to balance taste and energy consumption. The drying stress load index is used to describe the degree of influence of moisture gradient and temperature difference on grain structure during the drying process, and it is also a dimensionless indicator.

[0090] The process calculation unit identifies the set of operating time slices corresponding to each drying section in the thermal and moisture history fingerprint. Within each time slice, it estimates the moisture gradient and temperature difference between the inside and the surface of the grain based on the differences in moisture content and temperature at representative locations. Then, it compares these with the allowable safe gradient range for the variety. The safe gradient range is given in advance in the process specification based on the test results. For example, it can be specified that for a certain variety, when both the moisture gradient and temperature difference are in the lower half of the safe range, the drying stress load index is close to zero, and when the moisture gradient and temperature difference are close to the upper limit of the safe range, the drying stress load index is close to one.

[0091] The process calculation unit performs cumulative evaluation of the moisture gradient, temperature difference and corresponding residence time of each drying section on the time axis to obtain the drying stress load index of each drying section and the overall drying process. The typical value range is preferably zero to one. When the drying stress load index of a certain drying section is greater than 0.9 in three consecutive operating time segments, it can be regarded as the section approaching the crack risk boundary in the process specification, and the drying intensity of the section needs to be reduced first in subsequent steps.

[0092] After obtaining the cooking saturation index and drying stress load index of the current batch within a certain operating time segment or observation window, the process calculation unit compares them with the target finished product moisture content, brown rice hardness, and color index of the corresponding batch to form a quality offset for subsequent control.

[0093] The quality deviation is a vector reflecting the degree of deviation of the current operating state from the target quality. It can include components such as moisture content deviation, predicted hardness deviation, and predicted color deviation. The moisture content deviation can be obtained by extending the estimation of the subsequent drying process based on the thermal and moisture history fingerprint to obtain the predicted finished product moisture content, and then comparing it with the target finished product moisture content range. The predicted hardness deviation can be given as an empirical rule in the process specification or process calculation unit based on the correlation between the cooking saturation index and the drying stress load index and the actual hardness in historical batches. For example, it can be specified that when the cooking saturation index is lower than the target lower limit and the drying stress load index is higher than the target upper limit, the predicted hardness deviation is "too hard", and when the cooking saturation index is higher than the target upper limit and the drying stress load index is moderate, the predicted hardness deviation is "too soft" or "moderate". The predicted color deviation can be determined by summarizing the relationship between the temperature level and residence time experienced by the grain in the thermal and moisture history fingerprint and the color grade of historical batches to form an empirical rule, judging whether the current batch tends to be lighter or darker in color.

[0094] For each component of the quality deviation, the process specification can provide the target range and allowable deviation corresponding to the target quality of the variety. For example, the target range for moisture content deviation can be set as a symmetrical small range near zero, preferably set as a range from minus 0.5 percentage points to plus 0.5 percentage points. Hardness deviation is allowed to fluctuate within one grade near "moderate". Color deviation is allowed to fluctuate within the range of two adjacent color grades.

[0095] To avoid oversensitivity to short-term fluctuations in a single time slice, the process calculation unit statistically summarizes the quality offset within an observation window consisting of multiple running time slices. The observation window can be set to five to twenty consecutive running time slices, corresponding to several minutes to more than ten minutes of production time. Within this window, the quality offset is smoothed and trend analyzed. For example, the average offset value of each component within the observation window and the direction of offset change in the most recent three to five running time slices are calculated. When the average offset value within the observation window is still within the target range but the trend indicates that the deviation is continuously increasing, the process calculation unit can consider that the batch is deviating in an unfavorable direction.

[0096] Within the observation window, when the deviation trend of one or more components of the mass offset exceeds the preset threshold, the preset threshold can be defined in the process specification as: the moisture content deviation exceeds half of the target range, the predicted hardness deviation is "too hard" or "too soft" for more than half of the time slices within the observation window, the predicted color deviation indicates that the color is too dark or too light for more than half of the time slices within the observation window, etc. The process calculation unit combines the current cooking process and the remaining drying time, and calculates the stress quota and energy consumption quota of the remaining cooking and drying process of this batch according to the equipment safety process window and energy consumption budget.

[0097] The equipment safety process window refers to the pre-set safe operating range for each piece of equipment and each process parameter, including the maximum and minimum values ​​of steam temperature, air inlet temperature of each drying section, air volume, conveying speed, etc., as well as the upper limit of the rate of change. These ranges are given in the equipment manual and process specifications. The energy consumption budget refers to the range of steam and electricity that can be consumed for a batch under the premise of ensuring product quality. It can be given by the enterprise based on energy consumption targets and historical statistics.

[0098] Stress quota refers to the allowable drying stress and temperature intensity that can be allocated to each stage during the subsequent remaining cooking and drying processes, while energy consumption quota refers to the allowable increase or decrease in energy consumption during the remaining processes.

[0099] When calculating quotas, the process calculation unit determines the range of temperature, air volume, holding time, and conveying speed that can be increased or decreased in each drying and cooking section, without exceeding the equipment safety process window and overall energy consumption budget. For example, in a preferred embodiment, when the cooking saturation index of a certain batch is below 0.7 and the drying stress load index is below 0.5, the observation window length can be set to ten operating time slices, equivalent to about ten minutes. The process calculation unit can then allocate part of the stress quota to the cooking section in subsequent time slices, increasing the steam temperature by one to five degrees Celsius. Alternatively, the steaming and holding time can be extended by 5% to 20% within a safe range, while allocating some energy consumption quota to the subsequent low-temperature drying section to appropriately extend the low-temperature drying time; when the drying stress load index of a certain drying section is greater than 0.9 in three consecutive operating time segments, the process calculation unit can force the allocation of a negative stress quota to that section in the quota calculation, and reduce the upper limit of the allowable air inlet temperature and the upper limit of the air volume of that section by a preset range accordingly, and allocate the remaining drying tasks to other drying sections or complete them by increasing the overall drying time within the equipment safety process window.

[0100] When performing the above quota calculation, the process calculation unit will also check the status of the batch identifier in the batch library. If it finds that a batch identifier does not exist in the batch library or has been marked as ended, it will no longer calculate new stress quota and energy consumption quota for that batch, and can internally record error code E04. Error code E04 is used to indicate that the batch corresponding to the current quota request is in an invalid state, so as to avoid generating control policies for batches that have already ended.

[0101] In addition to the above-mentioned implementation method of quota calculation based on cooking saturation index and drying stress load index, an alternative implementation method can be adopted. In this method, the process calculation unit compares the thermal and moisture history fingerprint with the final quality results of historical batches during the operating condition analysis stage, and forms a comprehensive quality scoring rule through inductive analysis. The comprehensive quality score is a single dimensionless index that reflects the current comprehensive quality level of the thermal and moisture history. It can be set to a range of zero to one. The closer the value is to one, the closer the expected quality is to the target.

[0102] During operation, the process calculation unit calculates the comprehensive quality score according to the rule based on the thermal and moisture history fingerprint of the current batch, and compares the score with the target score range. The target score range can preferably be set to 0.8 to 1. When the average value of the comprehensive quality score within the observation window is lower than the lower limit of the target score range, the quality trend of the batch is considered unfavorable. The stress quota and energy consumption quota of each segment are adjusted through the same quota calculation logic as described above. As long as the deviation judgment and quota allocation logic corresponding to the comprehensive quality score are consistent with the aforementioned implementation method in principle, that is, when the score is low due to insufficient cooking, the quota of cooking and low-temperature drying is increased, and when the score is low due to excessive drying, the stress quota and energy consumption quota of the high-temperature segment are reduced, it can also be regarded as an equivalent implementation of this step.

[0103] Based on the above description, those skilled in the art can, within the actual production line's technological capabilities and applicable scope, and in conjunction with specific product varieties and equipment parameters, set reasonable index ranges, observation window lengths, and quota calculation rules, thereby reproducing the calculation process of this step and achieving the same technical effect without engaging in creative labor.

[0104] S5. Based on stress and energy consumption quotas, the steam temperature, steam pressure, and holding time in the cooking section, as well as the inlet air temperature, air volume, and conveying speed in each drying section, are adjusted in a coordinated manner to ensure that the thermal humidity history fingerprint follows the thermal humidity history template. Specifically, this is implemented as follows:

[0105] Based on the stress and energy consumption quotas determined for the current batch in step S4, the process calculation unit refines these quotas into adjustments to the actual equipment process parameters, so that the operating status of the cooking section and each drying section fits as closely as possible to the target trajectory corresponding to the selected heat and humidity history template for this batch on the time axis.

[0106] Therefore, during the system deployment phase, the equipment parameter library is used to register the safety process window and allowable variation range for each process parameter, such as steam temperature, steam pressure, cooking holding time of the cooking section, and air inlet temperature, air volume, and conveying speed of each drying section. The equipment safety process window refers to the parameter value range that has been confirmed to be acceptable for both equipment safety and product quality during long-term operation verification. For example, the steam temperature of a certain cooking equipment can be allowed to vary within the range of 90 degrees Celsius to 120 degrees Celsius, the air inlet temperature of each drying section can be allowed to vary within the range of 70 degrees Celsius to 130 degrees Celsius, and the conveying speed can be allowed to vary between the minimum and maximum speeds given in the equipment specifications. These minimum and maximum speeds are registered in the equipment parameter library with specific values. The allowable variation range is used to constrain the magnitude and rhythm of a single adjustment action. For example, the magnitude of a single adjustment of the air inlet temperature can be set to no more than 5 degrees Celsius, the magnitude of a single adjustment of the conveying speed can be set to no more than 5% to 20% of the current value, and the upper limit of the rate of change can be set to allow a maximum of three to five small adjustments to the same parameter within an observation window to avoid sudden shocks to the equipment and products.

[0107] During operation, when the mass deviation reaches the adjustment threshold preset in step S4 within the observation window, the process calculation unit, under the constraints of the stress quota and energy consumption quota of the current batch, calculates the target changes in steam temperature, steam pressure, cooking holding time, and air inlet temperature, air volume, and conveying speed of each drying section, based on the safety process window and allowable variation range registered in the equipment parameter library. The target changes are a set of parameter increments that cause the thermal humidity history fingerprint to deviate from the thermal humidity history template and return to the target range without exceeding the safety process window and allowable variation range.

[0108] When the cooking saturation index is below the lower limit of the target range, the process calculation unit prioritizes allocating stress and energy quotas to the cooking section, focusing the target changes on increasing steam temperature and appropriately extending cooking holding time, while constraining the changes in each drying section to prevent exacerbating drying stress. When the drying stress load index of a certain drying section approaches the risk limit, the process calculation unit prioritizes allocating negative stress quotas to that section, reducing the inlet air temperature or air volume of that section while maintaining the overall energy consumption budget balance, and extending the material residence time of that section within the safe process window. Through this kind of coordinated adjustment, the instantaneous drying stress of that section is mitigated, and the deviation in the thermal humidity history fingerprint is gradually pulled back to the range of the thermal humidity history template.

[0109] After calculating the target changes for each parameter, the process calculation unit adds these changes to the current parameter values ​​to form new target parameter values, and then organizes them into control command messages to send to the field control equipment.

[0110] A control command message is a set of control information sent from the process calculation unit to the field control equipment. Preferably, it includes fields such as command sequence number, batch identifier, section number, target steam temperature, target steam pressure, target cooking holding time, target inlet air temperature, target air volume, target conveying speed, planned execution time, and strategy version number. The strategy version number is used to identify which version of stress quota and energy consumption quota calculation rules and thermal humidity history template version the current command is based on.

[0111] Control command messages are sent between the process calculation unit and various field control devices through the aforementioned industrial network. The industrial network can be a layered structure of fieldbus processing industrial Ethernet, with the upper layer used to transmit batch information, command messages and operation results, and the lower layer used to transmit measurement signals.

[0112] To ensure the idempotency and sequentiality of control actions, the system assigns a unique instruction sequence number to each control instruction message across the entire system. The instruction sequence number can be generated by combining a timestamp and a sequence number. The field control device records the sequence number of the most recently successfully executed instruction and the corresponding strategy version number in its local non-volatile storage. When a control instruction message with the same sequence number recorded locally is received again, the message is considered a duplicate message, and only the time of receipt is recorded in the log. The process parameters are not adjusted again, thus avoiding duplicate execution caused by message retransmission or network jitter.

[0113] To ensure the sequential execution of instructions, the field control equipment checks the instruction sequence number continuously each time it receives a new message. The instruction sequence number can be required to monotonically increase relative to the most recently successfully executed instruction sequence number. If a jump in the instruction sequence number is detected—that is, the new instruction sequence number is greater than the most recently successfully executed sequence number plus the maximum allowable step size preset in the process specification, or a duplicate instruction is found that is smaller than the most recently successfully executed sequence number—the field control equipment can immediately send feedback information containing error code E02 to the process calculation unit via the industrial network. Error code E02 indicates an abnormal instruction sequence number, prompting the upper-level system to confirm or resend the instruction. Simultaneously, the field control equipment maintains the current process parameters unchanged to avoid performing potentially erroneous adjustments when the instruction sequence is uncertain.

[0114] To further ensure equipment operation safety, when a target parameter value in a control command message exceeds the equipment's safe process window—for example, if the control command requires an inlet air temperature higher than the maximum allowable temperature of the drying section or a conveying speed lower than the minimum allowable speed—the field control equipment performs a boundary check on the target value locally. The excess portion is truncated to the upper or lower limits of the safe process window registered in the equipment parameter library, and an adjustment is performed according to the truncated parameter value. Simultaneously, the next feedback message includes error code E03 along with the type and direction of the truncated parameter. Error code E03 indicates that the target value of this command has been adjusted by the safety mechanism, enabling the process calculation unit to identify the difference between the actual executed parameters and the issued target value in subsequent quota calculations.

[0115] In industrial networks experiencing momentary jitter or short-term retransmissions, to avoid performing repeated actions on the same equipment within a very short timeframe, field control equipment can deduplicate multiple messages with identical content and the same instruction sequence number within a set short time window. This short time window can be set to, for example, one to five seconds. Within this window, parameter adjustments are only performed on the first valid message. Subsequent duplicate messages are simply recorded for reception without being executed again, thereby reducing unnecessary adjustment operations while ensuring responsiveness.

[0116] By employing arrangements such as instruction sequence number uniqueness, local recording of the most recently successfully executed sequence number, sequence number continuity verification, and short time window deduplication, the adjustment of process parameters is made idempotent and ordered at the system level, avoiding duplicate adjustments and out-of-order execution caused by communication anomalies.

[0117] In a preferred embodiment, taking the production of a certain parboiled rice variety as an example, when the drying stress load index calculated in step S4 shows that the drying stress load index of the first drying section is greater than 0.9 in three consecutive operating time slots, the process calculation unit can allocate a negative stress quota to the first drying section in the stress quota and energy consumption quota of this batch, and generate a control command message accordingly, reducing the target inlet air temperature of the section by 5 to 10 degrees Celsius relative to the current value, and increasing the conveying residence time by 5 to 10 seconds. The control command message contains the new target inlet air temperature and the new conveying speed parameters. After the field control equipment completes the safety check and performs the adjustment, the thermal and humidity history fingerprint formed in the subsequent operating time slots will show the reduction of the moisture gradient and temperature difference of the section, and the drying stress load index will gradually fall back from the range close to 1 to the safe range. The thermal and humidity history fingerprint as a whole returns to the target trajectory range given by the corresponding thermal and humidity history template.

[0118] Based on the above description, those skilled in the art can configure instruction sequence numbers, error codes, and control instruction message structures within the actual production line's technological capabilities and applicable scope, combined with the specific equipment's safety process window and network conditions. This enables the coordinated adjustment of parameters in the cooking section and each drying section, ensuring the traceability, idempotency, and sequentiality of control actions. Consequently, the computational process of this step can be reproduced, and the same technical effect can be obtained.

[0119] S6. Record the feature vectors, thermal and humidity history fingerprints, control parameters, and final quality results for each batch. Periodically correct the thermal and humidity history model and thermal and humidity history template parameters to form an integrated control strategy that learns across batches. The specific implementation is as follows:

[0120] In order to form a traceable, optimizable, and self-learning control loop across batches on the actual integrated parboiler cooking and drying production line, the production line centrally records and versions key information of each batch in the process calculation unit or enterprise-level process management platform.

[0121] Each batch has a batch feature vector generated in step S1 when it enters production. In steps S3, S4 and S5, a thermal and humidity history fingerprint sequence and process parameters and control actions at the running time slice level are formed for that batch. In step S6, this information and the final quality results are summarized into a complete batch record.

[0122] The final quality results include the moisture content, brown rice hardness, color grade, and grade rice yield of the batch of finished products. Grade rice yield refers to the proportion of parboiled rice that reaches the target grade to the total quality of the batch of finished products. It is usually formed by the sorting equipment and quality inspection process. Moisture content and brown rice hardness can be given by laboratory testing or online testing devices, and color grade can be given by standard color plate comparison or image analysis system.

[0123] After each batch is completed, the process calculation unit organizes the batch feature vector, the thermal and moisture history fingerprint of the batch across all running time slices, the process parameters effective in each time slice, the corresponding control instruction execution records, and the aforementioned final quality results into one or more structured records and writes them into the running database and historical batch database. The structured records are stored in a field-separated format. The fields include batch identifier, variety code, initial moisture content of raw grain, soaking parameters, thermal and moisture history template version number, thermal and moisture history fingerprint summary, representative process parameters of key running time slices, statistical characteristics of cooking saturation index and drying stress load index on the time axis, final finished product moisture content, brown rice hardness, color grade, and grade rice yield, etc., and also includes the recording time, equipment number involved in production, process rule version number, and control strategy version number.

[0124] The process rule version number is used to identify the process procedure and judgment threshold version used when the batch is executed, and the control strategy version number is used to identify the quota calculation and parameter adjustment strategy version used when the batch is executed. Together with the heat and humidity history model version number and the heat and humidity history template version number, they constitute the version information of the evidence chain, which enables the complete working conditions, control behaviors and corresponding rule boundaries of the batch from raw grain entering the production line to finished product exiting the production line to be restored in the historical batch database based on the batch identifier during batch quality traceability, internal audit or external supervision inspection.

[0125] In this implementation, the chain of evidence refers to a set of associated records established around each batch, consisting of batch feature vectors, process records, control strategy versions, and final quality results. Once formed, it is not overwritten and can only be supplemented through incremental records. It can also be protected by read-only access permissions to ensure objectivity during tracing.

[0126] During the long-term operation of the production line, process engineers can conduct centralized analysis of the evidence chain records within a preset time period. This time period can be set to one week, two weeks, or one month to balance the amount of data and the workload of analysis. At the end of each period, the batch records of each variety, each source of raw grain, and each season are summarized and compared.

[0127] Process engineers can preferably group batch feature vectors of a certain variety under different seasons or different supplier conditions, compare the thermal and moisture history fingerprints in each group with the final quality results, identify thermal and moisture history trajectories that can stably obtain higher grade rice yield, qualified moisture content and target hardness under given raw grain conditions and have a low proportion of cracked batches, regard the model parameters and template parameters corresponding to such trajectories as representative values ​​of stable operating conditions, and adjust the parameters of the integrated cooking and drying thermal and moisture history model and thermal and moisture history template in an offline environment.

[0128] During the adjustment process, process engineers can fine-tune the heat transfer and moisture migration parameters in the model by combining data from multiple batches. This reduces the deviation between the model's predicted thermal and moisture history and the actual measured thermal and moisture history fingerprint. The target temperature and moisture content trajectory bands in the template can be appropriately tightened or widened to better reflect the actual achievable stable operating conditions. After the adjustment is complete, the new model parameters and template parameters are registered as new versions in the model and template libraries, assigned new version numbers and effective times, and a version locking mechanism is enabled in the version management strategy. This means that the new version only applies to batches started after the version switch time. Batches already in production and running before the version switch continue to use the original version of the model and template. This avoids changes to the thermal and moisture history target curve and quota calculation logic midway through a batch, which could cause control behavior jumps and product quality fluctuations.

[0129] In addition to adjusting the model and template parameters, the system can also evaluate the overall performance of the current integrated control strategy based on the grade yield, unit product energy consumption, and the proportion of batches with insufficient gelatinization or excessive cracks recorded in the historical batch database.

[0130] Unit product energy consumption can be expressed as the steam consumption and electrical energy consumption corresponding to a unit mass of parboiled rice, provided by energy consumption metering devices and production metering systems; insufficient gelatinization batches can be determined by the cooking saturation index being consistently below the lower limit of the target range and the hardness being relatively high; batches with excessive cracks can be determined by the drying stress load index being consistently high and the proportion of cracked particles exceeding the enterprise standard during inspection.

[0131] The system can periodically run evaluation programs within an enterprise-level process management platform or process calculation unit. When it is found that the yield of a certain product grade is lower than the target level predetermined in the process specification in each evaluation period for the past two to five evaluation periods, or the energy consumption per unit product is higher than the upper limit of the target range for the past two to five evaluation periods, or the proportion of batches with insufficient gelatinization and the proportion of batches with excessive cracks continuously exceed the allowable range specified in the enterprise specification for the past two to five evaluation periods, the corresponding batches and corresponding operating conditions will be marked as key analysis objects. The process engineer will be reminded to adjust the relevant thermal and humidity history templates, stress quota rules and energy consumption budgets through internal notifications or interface prompts.

[0132] Through this evidence-based periodic analysis and parameter updates, the integrated control strategy can gradually adapt to the actual production environment under conditions of cross-seasonal, cross-source of raw materials and cross-equipment status changes, forming a long-cycle optimization mechanism with self-learning capabilities.

[0133] In a preferred embodiment, under the conditions of a certain main variety and relatively stable raw grains, after adopting this technical solution, while keeping the equipment configuration unchanged, compared with an integrated production line operating with fixed process parameters, after one to two evaluation cycles of model and template adjustment, the yield of graded rice can be increased by two to five percentage points, the unit steam consumption and unit electricity consumption can be reduced by five to fifteen percent respectively, and the proportion of batches with excessive cracks will be significantly reduced.

[0134] The process calculation unit in this step can be deployed as a distributed control system, industrial computer, or edge computing device. Alternatively, it can be completed collaboratively by the field control system and the cloud-based process platform. The collaboration can be achieved by the field system completing the thermal and humidity history fingerprint calculation and control command issuance within the runtime slice, while the cloud platform completes the batch record archiving, statistical analysis, and model template version update within the evaluation period. As long as the closed-loop delay from the end of the runtime slice to the issuance of the corresponding control decision does not exceed half the runtime slice length and meets the enterprise's requirements for production continuity and information security, it can be considered an equivalent implementation of "recording the feature vectors, thermal and humidity history fingerprints, control parameters, and final quality results of each batch, periodically correcting the thermal and humidity history model and thermal and humidity history template parameters, and forming an integrated control strategy with cross-batch self-learning."

[0135] Based on the above description, those skilled in the art can establish a historical batch library, evidence chain management mechanism, and version locking strategy within the actual production line's process capabilities and applicable scope. This step can be combined with the aforementioned steps S1 to S5 to achieve a complete technical solution from single-batch closed-loop control to cross-batch self-learning optimization.

[0136] In one operating scenario shown in this embodiment: On a parboiled rice cooking and drying integrated production line with a rated processing capacity of 20 tons per hour and equipped with a single-stage cooking and three-stage co-current drying equipment, after the enterprise has deployed the process calculation unit, on-site control equipment, batch library, operation library, model library and template library according to the aforementioned steps S1 to S6, it begins continuous production of a certain main rice variety. When the morning shift begins, a batch of raw grain from warehouse A01, with supplier code S01, is ready to enter the production line. Before the raw grain enters the pre-processing section, the system automatically generates a batch identifier containing the date, time, warehouse number, and supplier code, such as "20260109-083000-A01-S01". Under the condition that the online moisture measuring device above the feed hopper is operating stably, the moisture content of this batch of raw grain is detected during the process of entering the production line. At this time, the initial moisture content is found to be around 14.5% by mass. This batch of raw grain is identified as a medium-grain size parboiled grain variety in the enterprise coding table, with the corresponding variety code "V01". The on-site online screening device sampling results show that the particle size grade is the medium particle size grade in the enterprise standard. The soaking tank control device shows that this batch of raw grain has been soaked in warm water at about 50 degrees Celsius for about five hours. The soaking time and soaking water temperature are recorded and fixed by the timing device and temperature sensor at the beginning and end of the soaking.

[0137] The initial moisture content, variety code, particle size grade, and soaking parameters mentioned above are organized by the process calculation unit into a batch feature vector for this batch. In accordance with the process specifications for variety "V01", the target finished product moisture content is bound to the range of 13% to 13.5% by mass fraction, the target brown rice hardness is bound to the mechanical test range corresponding to the suitable taste of this variety, and the color index is bound to the target grade of parboiled rice color grade within the enterprise. These target indicators and the batch feature vector are written into the batch database together, along with the process rule version number and the recording time, forming the initial evidence chain record for this batch.

[0138] Since the production line had already established and solidified the integrated cooking and drying thermal process model and thermal process template in the previous stage according to step S2, when the batch feature vector enters the process calculation unit, the system searches the template library based on the variety code "V01" and the target finished product moisture content, brown rice hardness, and color indicators. It selects a thermal process template that perfectly matches "V01" and is suitable for the current target quality combination. This template requires maintaining the internal temperature of the grains between 90 and 100 degrees Celsius for approximately 25 minutes during the cooking stage. In the first drying stage, surface moisture is rapidly removed at an inlet air temperature of approximately 115 degrees Celsius. In the second drying stage, the overall moisture content is reduced at an inlet air temperature of approximately 95 degrees Celsius. In the third drying stage, a gentle drying process is performed at approximately 78 degrees Celsius, where the internal and external temperature difference is relatively small, ultimately reducing the internal moisture content of the grains to the target range. The template's identifier and version number, along with the batch identifier, are stored in the batch database and the batch additional information field in the runtime database as a target reference for subsequent runs of this batch.

[0139] As the batch of raw grain "20260109-083000-A01-S01" entered the cooking section after soaking, the on-site control equipment, under the unified time reference distributed by the process calculation unit, divided the continuous production time into sixty-second running time slices. Each running time slice was uniquely identified between the on-site control equipment and the process calculation unit through a time slice number plus a second-level timestamp. Within the first ten minutes, each on-site control equipment, as agreed, collected and aggregated the steam medium temperature of the cooking section, the temperature of a representative location inside the grain layer, the humidity of the cooking section, and the moisture content at the cooking outlet at the end of each running time slice, as well as the inlet air temperature, exhaust air temperature, exhaust air humidity, and outlet moisture content of the first drying section. For the second and third drying sections, since the material had not yet arrived, only temperature and humidity data were submitted.

[0140] The aggregated temperature, humidity, moisture content, and residence time are encapsulated together with the operating time slice number, equipment number, batch identifier, and segment number into a working condition transmission message. This message is transmitted to the process calculation unit via fieldbus and industrial Ethernet. The process calculation unit aligns and cleans these data in the runtime library according to the time slice, removes occasional out-of-bounds values, and interpolates the few missing moisture values ​​based on the trend of the preceding and following time slices. Then, it calls the aforementioned integrated cooking and drying thermal and moisture history model to update the status of representative positions in the grain layer in space with the time slice as the step size. This yields the temperature and moisture content of each position inside the grain at the end of each time slice and compares it with the thermal and moisture history template selected for that batch, thereby forming the thermal and moisture history fingerprint sequence of that batch in the runtime library.

[0141] As the material gradually passes through the second and third drying sections, the process calculation unit calculates the corresponding cooking saturation index and the drying stress load index of each drying section within an observation window consisting of ten operating time slices, based on the thermal and moisture history fingerprint. In this operating scenario, due to the stable steam supply in the early stage, the cooking saturation index rises rapidly to above 0.7, approaching the target range. However, after one hour of continuous production, the external heating conditions fluctuate slightly, and the air inlet temperature of the first drying section is slightly higher than the template's recommended value. This causes the drying stress load index of this section to approach 1 in some time slices within the three consecutive observation windows. At this time, the comparison between the thermal and moisture history fingerprint in the operating database and the template indicates that the surface moisture of the grains in this section decreases rapidly while the internal moisture gradient is relatively large. Based on this, the process calculation unit judges that this batch has a tendency to approach the crack risk boundary in the first drying section.

[0142] Within the same observation window, the process calculation unit estimates the predicted finished product moisture content and predicted hardness under the current trend by extending the thermal and moisture history fingerprint. Combined with the target finished product moisture content and target hardness of this batch, it calculates the quality deviation and finds that the moisture content deviation is still within the allowable range, while the predicted hardness shows a trend towards being harder. According to the threshold conditions preset in the process specification, the process calculation unit considers the quality trend of this batch to be unfavorable, triggering the redistribution logic of stress quota and energy consumption quota. Under the premise of not exceeding the safety process windows of the cooking equipment and each drying section, and ensuring that the overall energy consumption budget of this batch does not exceed the limit, a portion of the stress quota is recovered from the first drying section and transferred to the subsequent low-temperature drying section, while a small amount of usable quota is reserved for the cooking section.

[0143] Subsequently, the process calculation unit generates new target changes for the target process parameters of the first and third drying sections based on the quota calculation results, forming control command messages. The target change for the first drying section is to reduce the inlet air temperature by approximately eight degrees Celsius relative to the current value and slightly reduce the air volume within a safe range. The target change for the third drying section is to moderately extend the material residence time. For the cooking section, the steam temperature will not be increased but will only be maintained at the current level. The control command message carries a unique command sequence number, batch identifier, section number, new target temperature, air volume and conveying speed, as well as the planned execution time and strategy version number, and is sent to the corresponding field control equipment via the industrial network.

[0144] After receiving the control command message, the on-site control equipment performs boundary checks on the target values ​​according to the safety process window registered in the equipment parameter library. It confirms that the new target inlet air temperature and air volume are within the allowable range. Subsequently, the local control logic adjusts the process parameters of the first and third drying sections at the planned execution time. In the subsequent multiple operating time slices, the operating condition transmission messages and thermal humidity history fingerprints show that the moisture gradient and temperature difference of the first drying section gradually decrease, and the drying stress load index drops from close to one to about 0.7. Meanwhile, the moisture content curve inside the third drying section becomes flatter as it approaches the finished product section. Finally, the predicted hardness deviation of this batch returns to the target range.

[0145] As batch "20260109-083000-A01-S01" completed its final drying stage and entered the cooling and sorting section, the laboratory randomly sampled and tested the batch. The resulting finished product had a moisture content concentrated at approximately 13.2% by mass, brown rice hardness within the middle of the target range allowed by the enterprise standard, color indicators met the predetermined grade, and the graded rice yield was approximately 96%. The unit steam and electricity consumption was slightly lower than the historical average when operating with fixed process parameters. At the end of the batch, the process calculation unit compiled a complete batch evidence chain, including the thermal and moisture history fingerprint summary of the batch across all operating time segments, the statistical characteristics of key process parameters for each time segment, cooking saturation index, and drying stress load index, as well as the final quality results, graded rice yield, unit energy consumption indicators, and error code records and control command execution records that occurred during operation. This chain was then stored in the historical batch database.

[0146] In the subsequent one or two evaluation cycles, process engineers compared the thermal and moisture history fingerprints and quality results of variety "V01" under different seasons and different suppliers based on evidence chains from multiple similar batches on the enterprise-level process management platform. They found that the current thermal and moisture history template and quota rules could stably maintain a high grade rice yield and a low crack ratio under the conditions of this type of raw grain. Only minor adjustments to the temperature trajectory band at the end of the third drying stage were needed to further optimize energy consumption performance. Therefore, a new template version was formed and a version locking strategy was adopted to gradually promote it in subsequent new batches.

[0147] All calculations involved in the embodiments are dimensionless numerical calculations, and the preset parameters and thresholds in the calculations are set by those skilled in the art according to the actual situation.

[0148] It should be noted that this invention can be deployed on the device itself to realize embedded applications, or it can run on a PC or other terminal with a user interface, thereby meeting various hardware environments and usage requirements.

[0149] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wireless or wired transmission; wired transmission methods include optical fiber, twisted pair, coaxial cable, etc.; wireless transmission includes infrared, microwave, etc. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center containing one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0150] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and modules described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0151] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.

[0152] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0153] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0154] If the aforementioned functions are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0155] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0156] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for integrated control of parboiled rice cooking and drying, characterized in that, include: S1. Collect the initial moisture content, variety, particle size and soaking parameters of raw grains in batches, generate batch feature vectors and bind them to the target finished product moisture content, brown rice hardness and color index; S2. Based on the batch feature vector and the process parameters of the cooking section and each drying section, establish an integrated heat and humidity process model for cooking and drying, generate a heat and humidity process template that meets the target indicators offline, and solidify the model parameters. S3. Divide the operation into time slices under unified time synchronization, collect the temperature, humidity, moisture content and residence time of the cooking section and each drying section according to the time slice, and use the thermal and humidity history model to obtain the thermal and humidity history fingerprint of this batch. S4. Calculate the cooking saturation index and drying stress load index based on the thermal and wet history fingerprint, combine them with the target index to form the mass offset, and determine the stress quota and energy consumption quota for the remaining cooking and drying processes. S5. Based on the stress quota and energy consumption quota, the steam temperature, steam pressure, holding time of the cooking section, as well as the air inlet temperature, air volume, and conveying speed of each drying section are adjusted in a coordinated manner so that the heat and humidity process fingerprint follows the heat and humidity process template. S6. Record the feature vectors, thermal and humidity history fingerprints, control parameters and final quality results of each batch, periodically correct the thermal and humidity history model and thermal and humidity history template parameters, and form an integrated control strategy for cross-batch self-learning.

2. The integrated control method for parboiled rice cooking and drying according to claim 1, characterized in that, S1 includes: A batch identifier consisting of date, time, warehouse number, and supplier code is generated for each batch of raw grain entering the production line; The initial moisture content, variety code, particle size grade, soaking time and soaking water temperature of the batch of raw grains are collected and used to construct a batch feature vector according to preset fields and coding rules. According to the variety process specifications, the target finished product moisture content, brown rice hardness and color index are bound to this batch, and the process calculation unit writes the batch feature vector and target index along with the process rule version number into the batch database.

3. The integrated control method for parboiled rice cooking and drying according to claim 1, characterized in that, S2 include: The process calculation unit establishes an integrated heat and humidity process model for cooking and drying based on batch feature vectors, cooking section process parameters, and drying section process parameters. The grain layer thickness was divided into multiple representative locations, the running time was divided into fixed-length time slices, the model parameters were adjusted and the model version number was set according to the working condition records. Based on the target finished product's moisture content, brown rice hardness, and color indicators, a heat and humidity history template is generated. The correspondence between the template identifier and the variety code and model version number is established, and the template identifier is stored in the template library.

4. The integrated control method for parboiled rice cooking and drying according to claim 1, characterized in that, S3 include: Under a unified time synchronization standard, continuous production time is divided into run time slices with run time slice numbers; The on-site control equipment collects the temperature, humidity, moisture content and residence time of the cooking section and the drying section during each operating time segment, and generates operating condition transmission messages carrying batch identifiers and section numbers. The process calculation unit aligns and cleans the operating condition reports according to the running time slice number, calls the integrated cooking and drying heat and moisture history model to estimate the temperature and moisture content of representative locations, and organizes the estimation results into heat and moisture history fingerprints and writes them into the running library.

5. The integrated control method for parboiled rice cooking and drying according to claim 1, characterized in that, S4 include: Based on the thermal and moisture history fingerprint in the runtime library, the process calculation unit identifies the set of running time slices corresponding to the cooking section and each drying section. Within each running time slice, it calculates the cooking saturation index and drying stress load index based on the temperature and moisture content of representative locations. The cooking saturation index and drying stress load index are then compared with the target finished product moisture content, brown rice hardness, and color index of the batch to generate a quality offset.

6. The integrated control method for parboiled rice cooking and drying according to claim 5, characterized in that: The process calculation unit performs statistical analysis on the mass offset within the observation window consisting of continuous running time slices and identifies the trend of mass offset change. Based on the degree of deviation of the cooking saturation index and drying stress load index from the target range, combined with the equipment safety process window and energy consumption budget, it determines the stress quota and energy consumption quota of the remaining cooking section and each drying section of this batch.

7. The integrated control method for parboiled rice cooking and drying according to claim 1, characterized in that, S5 include: The process calculation unit registers the equipment safety process window and allowable variation range for the process parameters of the cooking section and drying section in the equipment parameter library; When the mass offset reaches the adjustment threshold, the parameter change is calculated under the constraints of stress quota and energy consumption quota, and an instruction message containing instruction number, batch identifier, parameters and strategy version number is generated and sent to the field control equipment through the industrial network. The field control equipment identifies duplicate messages and sequence number abnormalities based on the recorded execution instruction sequence number. When the parameters exceed the equipment's safe process window, it performs boundary checks and generates error code feedback information.

8. The integrated control method for parboiled rice cooking and drying according to claim 1, characterized in that, S6 include: After each batch is completed, the process calculation unit writes the batch feature vector, thermal and humidity history fingerprint, key operating time slice process parameters, control instruction execution records and final quality results of that batch into the runtime library and historical batch library in a field-based record format. The batch identifier, thermal and humidity history model version number, thermal and humidity history template version number, process rule version number, and control strategy version number are written into the runtime library and historical batch library in a field-based manner to form the evidence chain record for that batch.

9. The integrated control method for parboiled rice cooking and drying according to claim 8, characterized in that: The process calculation unit groups batch records according to variety and raw grain conditions within a preset time period, performs statistical analysis on the thermal and moisture history fingerprints of each group and the final quality results, and determines the model parameters and thermal and moisture history template parameters corresponding to stable operating conditions. The updated model parameters and thermal-humidity history template parameters are registered in the model library and template library as new versions. At the same time, a version locking strategy is adopted to make the new version effective in the batch launched after the version switch, so as to achieve cross-batch self-learning optimization.

Images