A gradient temperature control system and method for automatic processing of tiger skin chicken feet

Abstract

This invention belongs to the field of food processing technology. It discloses an automated gradient temperature control system and method for processing tiger-skin chicken feet, comprising: real-time data acquisition; generating a multi-source fusion sensing set through three-level verification; performing multi-condition logic to determine the current cooling stage, and performing feedforward pre-compensation for the initial state of the environment and product, executing a dynamic priority strategy, and generating a gradient temperature control decision instruction set; driving the actions of each actuator, performing feedback compensation, and generating gradient temperature control execution results; determining cooling completion, generating a cooling report and sending it to the subsequent braising process, forming collaborative data and batch temperature control archives and archiving them in a historical database; and optimizing the gradient temperature control curve and feedforward pre-compensation parameters through multi-dimensional performance evaluation, updating the gradient temperature control decision instruction set as the initial judgment basis for the next round of multi-condition logic, forming a full-process optimization link.

Description

Technical Field

[0001] This invention relates to the field of food processing technology, and more specifically, to an automated gradient temperature control system and method for processing tiger-skin chicken feet. Background Technology

[0002] With the intelligent transformation of food processing, the demand for refined control of the cooling process in the large-scale production of tiger skin chicken feet has become prominent: cooling needs to take into account safety (rapidly passing through dangerous temperature zones), quality (controlling surface moisture and core temperature uniformity), efficiency (adapting to individual weight / environmental variables) and cross-process collaboration (connecting frying and braising). However, the current traditional methods that rely on fixed parameters or single sensors cannot meet these needs, resulting in safety risks, quality fluctuations and efficiency bottlenecks.

[0003] Existing solutions have several shortcomings in addressing current control requirements: they only collect temperature data in a single dimension, lack simultaneous acquisition and verification of multiple parameters, resulting in unreliable data; they use static temperature control curves, making it difficult to adapt to product or environmental variables; when safety requirements (such as rapid cooling) conflict with quality requirements (such as controlling surface moisture), the solution can only rely on temporary manual intervention when there is a conflict between safety requirements (such as rapid cooling) and quality requirements (such as controlling surface moisture); and the existing solutions isolate data from the cooling process and the processes before and after them, resulting in poor coordination. Summary of the Invention

[0004] To overcome the aforementioned deficiencies of the prior art and to achieve the above objectives, the present invention provides the following technical solution: an automated processing gradient temperature control system for tiger-skin chicken feet, comprising:

[0005] Multi-source data acquisition and fusion module: Real-time acquisition of data related to cooling medium, environment and products, combined with data from the previous frying process, and through three-level verification, generates a multi-source fusion sensing set;

[0006] Stage Judgment Decision Generation Module: Based on a multi-source fusion sensing set, the module determines the current cooling stage through multi-condition logic, calls the corresponding gradient temperature control curve, performs feedforward pre-compensation for the initial state of the environment and product, and combines batch risk identification to execute a dynamic priority strategy to generate a gradient temperature control decision instruction set.

[0007] Actuator drive compensation module: Based on the gradient temperature control decision instruction set, it drives the actions of each actuator, collects the core temperature and surface data of the product in real time for feedback compensation, and generates gradient temperature control execution results;

[0008] Cooling completion determination module: Based on the results of gradient temperature control, when the core temperature of the product meets the expectation, cooling is triggered, a cooling report is generated and sent to the subsequent brining process and feedback is received, forming collaborative data and batch temperature control files and archiving them to the historical database;

[0009] Iterative optimization module: Based on multiple batches of data from the historical database, it optimizes the gradient temperature control curve and feedforward pre-compensation parameters through multi-dimensional performance evaluation, updates the gradient temperature control decision instruction set, and serves as the initial judgment basis for the next round of multi-condition logic, forming a full-process optimization link.

[0010] Furthermore, the generation method of the multi-source fusion sensing set includes:

[0011] Collect multi-source data from the cooling process and categorize it into cooling medium data, environmental data, and product data according to data type; simultaneously, use RFID readers to read batch risk identifiers and key parameters from the preceding frying process and merge them with product data to form product-related data;

[0012] The three-level verification logic is defined as: Level 1 integrity verification, Level 2 consistency verification, and Level 3 stability verification.

[0013] All acquired data undergoes three levels of verification to identify invalid data and mark fluctuating data, thereby retaining valid data and marked fluctuating data, and integrating them into a structured multi-source fusion sensing set.

[0014] Furthermore, the method of determining the current cooling stage through multi-condition logic includes:

[0015] Based on a multi-source fusion sensing set, the product's average core temperature, cooling duration, and real-time cooling rate are extracted as core feature parameters.

[0016] The cooling stage is divided into three categories: pretreatment stage, rapid cooling stage, and temperature equalization stage.

[0017] The multi-condition logic is defined as follows: based on the preset stage threshold range of the feature parameters, the threshold of the feature parameters is determined through the AND logic of the logic gate to identify the current cooling stage;

[0018] If parameters overlap during the threshold determination process, the average core temperature of the product shall be the primary criterion for determination.

[0019] The current cooling stage is determined by multi-condition logic, and the determined stage type and corresponding feature parameters are packaged to generate a stage determination result package.

[0020] Furthermore, the method for performing feedforward pre-compensation for the initial state of the environment and product includes:

[0021] Based on the stage determination result package, according to the current cooling stage type, the corresponding stage's benchmark gradient temperature control curve is called from the preset stage-gradient curve mapping library; at the same time, environmental parameters and product initial parameters are extracted from the multi-source fusion sensing set.

[0022] Feedforward pre-compensation is defined as follows: by using environmental parameters and initial product parameters, and according to preset compensation rules, the reference gradient temperature control curve is corrected to generate a compensated gradient temperature control curve, and packaged into a gradient curve-compensation result package.

[0023] Furthermore, the generation method of the gradient temperature control decision instruction set includes:

[0024] Based on gradient curve-compensation result package and batch risk identifier, safety, quality and efficiency are taken as control objectives. First, the basic priority weights of the three control objectives are allocated according to the cooling stage. Then, the basic priority weights are fine-tuned in a targeted manner in combination with batch risk identifiers to obtain the adjusted priority weights.

[0025] Real-time verification of whether there are conflicts between the control objectives corresponding to the priority weights, and handling of existing control objective conflicts through the set general arbitration rules;

[0026] The arbitrated control target parameters are combined with the actuator characteristics and the parameters of the compensated gradient temperature control curve to transform them into general control commands for each actuator.

[0027] The common control instructions for all actuators are structured and integrated to generate a gradient temperature control decision instruction set.

[0028] Furthermore, the method of driving each actuator to move includes:

[0029] Based on the temperature control decision instruction set, the actuator type, control logic, triggering conditions, and control target priority of each instruction are extracted for each actuator.

[0030] Then, various actuators are driven to perform corresponding actions according to the priority order of the control target, while real-time data of the actuator's operating status is collected and recorded.

[0031] Furthermore, the method for generating the gradient temperature control execution result includes:

[0032] The core and surface data of the product are collected at high frequency, aligned with the actual operating status data by timestamp, and then compared with the compensated gradient temperature control curve to determine the parameter deviation.

[0033] Based on the determined deviation type, a targeted compensation instruction is generated according to the categorized compensation strategy to drive the actuator to adjust its actions;

[0034] After the compensation command is executed, the compensation effect is verified and anomalies are handled in a timely manner, generating gradient temperature control execution results that include execution process data and stage status data.

[0035] Furthermore, the methods for forming collaborative data and batch temperature control records and archiving them to the historical database include:

[0036] Based on the results of gradient temperature control, key judgment parameters are extracted, and judgment is performed according to the preset cooling completion standard. If the judgment is not completed, secondary compensation is triggered until the cooling is completed.

[0037] If the process is deemed complete, the cooling completion process is triggered, a cooling process report containing process traceability data and core result parameters is generated, and the report is sent to the subsequent brining process and parameter adaptation feedback is received.

[0038] Integrate cooling process reports and parameter adaptation feedback to form cross-process collaborative data and batch temperature control archives, which are then archived in a historical database.

[0039] Furthermore, the formation of the entire process link includes:

[0040] Extracting multiple batches of end-to-end data from historical databases to construct a four-dimensional evaluation mechanism:

[0041] S91: Safety, quality, efficiency and energy consumption are used as evaluation indicators, and the compliance rate of each evaluation indicator is defined.

[0042] S92: Calculate the individual score for each evaluation indicator according to the mapping rule of compliance rate → score, and obtain the batch comprehensive performance score by weighted summation. Select batches whose comprehensive scores do not meet expectations as key optimization batches.

[0043] S93: Conduct root cause analysis on key optimization batches to pinpoint performance shortcomings;

[0044] To address performance shortcomings, optimization schemes for gradient temperature control curves, feedforward pre-compensation parameters, and priority weights were developed based on simulation verification.

[0045] By updating the gradient temperature control decision instruction set through optimization and simultaneously using it as the initial judgment basis for the next round of multi-condition logic, a full-process optimization chain of data-driven, evaluation optimization, and closed-loop iteration is formed.

[0046] Furthermore, a gradient temperature control method for automated processing of tiger-skin chicken feet includes:

[0047] S1: Real-time collection of data related to cooling medium, environment and products, combined with data from the previous frying process, and generated a multi-source fusion sensing set through three-level verification;

[0048] S2: Based on a multi-source fusion sensing set, the current cooling stage is determined through multi-condition logic, the corresponding gradient temperature control curve is called, and feedforward pre-compensation is performed for the initial state of the environment and product. Combined with batch risk identification, a dynamic priority strategy is executed to generate a gradient temperature control decision instruction set.

[0049] S3: Based on the gradient temperature control decision instruction set, drive the actions of each actuator, collect the core temperature and surface data of the product in real time for feedback compensation, and generate gradient temperature control execution results.

[0050] S4: Based on the results of gradient temperature control, when the core temperature of the product meets the expectation, cooling is triggered to complete, a cooling report is generated and sent to the subsequent brining process and feedback is received, forming collaborative data and batch temperature control files and archiving them to the historical database;

[0051] S5: Based on multiple batches of data from the historical database, through multi-dimensional performance evaluation, optimize the gradient temperature control curve and feedforward pre-compensation parameters, update the gradient temperature control decision instruction set, and use it as the initial judgment basis for the next round of multi-condition logic, forming a full-process optimization link.

[0052] The technical effects and advantages of the automated gradient temperature control system and method for processing tiger-skin chicken feet according to the present invention are as follows:

[0053] This invention solves the problem of unreliable, single-dimensional data by synchronously collecting multi-dimensional data and performing three-level verification, thus providing an accurate basis for decision-making.

[0054] Secondly, by accurately determining the cooling stage, dynamically calling the curve and pre-compensating, it can also adjust the target priority and handle conflicts according to batch risk, breaking through the limitations of static strategies; then, it drives the actuator according to priority, collects data in real time and performs closed-loop compensation to solve the problem of unstable temperature control; then, by determining the cooling status, transmitting reports across processes and archiving data, it breaks the isolation of process data.

[0055] Finally, by optimizing parameters and updating the instruction set based on historical data, the system can be continuously upgraded, solving the problem that existing technologies cannot be continuously optimized. This allows the system to continuously improve its adaptability and control precision as production data accumulates, achieving an upgrade from stable operation to continuous optimization, and ensuring a comprehensive improvement in product safety, quality, and production efficiency in the long term. Attached Figure Description

[0056] Figure 1 This is a schematic diagram of an automated gradient temperature control system for processing tiger-skin chicken feet according to the present invention;

[0057] Figure 2 This is a schematic diagram of the cooling stage determination process in an automated gradient temperature control system for processing tiger skin chicken feet according to the present invention.

[0058] Figure 3 This is a schematic diagram of an automated gradient temperature control method for processing tiger-skin chicken feet according to the present invention. Detailed Implementation

[0059] 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.

[0060] Example 1

[0061] Please see Figure 1 and Figure 2 As shown in this embodiment, an automated gradient temperature control system for processing tiger-skin chicken feet includes:

[0062] Multi-source data acquisition and fusion module: Real-time acquisition of data related to cooling medium, environment and products, combined with data from the previous frying process, and generated a multi-source fusion sensing set through three-level verification.

[0063] It should be noted that this step provides the data foundation for subsequent cooling stage determination and gradient temperature control decisions.

[0064] Stage-based decision generation module: Based on a multi-source fusion sensing set, it determines the current cooling stage through multi-condition logic, calls the corresponding gradient temperature control curve, performs feedforward pre-compensation for the initial state of the environment and product, and combines batch risk identification to execute a dynamic priority strategy to generate a gradient temperature control decision instruction set.

[0065] It should be noted that this step provides a precise control basis for subsequent actuator actions, adapting to the current working conditions and risk level.

[0066] Actuator drive compensation module: Based on the gradient temperature control decision instruction set, it drives the actions of each actuator, collects the core temperature and surface data of the product in real time for feedback compensation, and generates gradient temperature control execution results.

[0067] It should be noted that this step achieves precise conversion from decision-making instructions to physical actions, as well as closed-loop control of execution-feedback-adjustment, thereby ensuring that the product status conforms to the target parameters and providing data support for subsequent cooling completion determination.

[0068] Cooling completion determination module: Based on the results of gradient temperature control, when the core temperature of the product meets the expectations, the cooling is triggered to complete, a cooling report is generated and sent to the subsequent braising process and feedback is received, forming collaborative data and batch temperature control files and archiving them to the historical database.

[0069] It should be noted that this step not only completes the closed-loop management of the current batch of cooling, but also provides complete data support for subsequent multi-dimensional performance evaluation and iterative optimization.

[0070] Iterative optimization module: Based on multiple batches of data from the historical database, it optimizes the gradient temperature control curve and feedforward pre-compensation parameters through multi-dimensional performance evaluation, updates the gradient temperature control decision instruction set, and serves as the initial judgment basis for the next round of multi-condition logic, forming a full-process optimization link.

[0071] It should be noted that this step, as the core iterative module of the system, not only solves the efficiency problem in current production, but also provides a more adaptable temperature control strategy for subsequent processes, realizing the technical upgrade of the gradient temperature control system from stable operation to continuous optimization.

[0072] The methods for generating multi-source fusion sensing sets include:

[0073] After the system starts up, it automatically performs an initialization self-test (sensor communication link detection, actuator response test, initial parameter loading). After the self-test passes, it triggers the data acquisition process, generates an acquisition start signal, and synchronously activates the multimodal sensor array and data interface.

[0074] Based on the acquisition start signal, multi-source data of the cooling process is collected according to the preset frequency and acquisition range. According to the data type, it can be divided into cooling medium data, environmental data and product data.

[0075] Cooling medium data: Real-time flow rate of the cooling medium (such as low-temperature water / cool air) is collected via an electromagnetic flow meter built into the pipeline (e.g., sampling frequency 5 seconds / time, recording unit m). 3 / h); collect medium temperature distribution through distributed temperature sensors (e.g., spaced 5cm apart) (e.g., sampling frequency 10 seconds / time, recording unit ℃), covering the entire area of ​​the cooling tank and pipeline;

[0076] Environmental data: The ambient temperature (range 0-40℃) and relative humidity (range 20%-90%) of the cooling area are collected by temperature and humidity transmitters deployed in the workshop, with a sampling frequency of 5 minutes / time.

[0077] Product data: The surface of the product (tiger skin chicken feet) is scanned by an infrared thermal imager to generate the surface temperature (the sampling frequency can be set to 10 seconds / time), the surface humidity (i.e., skin moisture) of the product is collected by a surface humidity detection sensor, and the core temperature of the product (i.e., the internal center temperature of the tiger skin chicken feet) is collected by a contact core temperature sensor.

[0078] Simultaneously, the RFID reader is triggered to read the carrier tag information and obtain key parameters of the previous frying process, including the outlet temperature (range 160-180℃) and the product weight; at the same time, the equipment carrier tag is read to obtain the batch risk label preset by the previous process (this label is usually classified by the previous process according to the raw material status, such as the freshness of the raw material, the fluctuation range of the process, etc., and the general level is divided into high risk and normal).

[0079] The collected data undergoes a three-level verification process, defined as follows: Level 1 integrity verification, Level 2 consistency verification, and Level 3 stability verification.

[0080] The first-level integrity check is as follows: check whether there is any missing data from each sensor (for example, if there is no return value after 3 consecutive samplings, it is determined to be missing data). If the missing data is marked as invalid data, if there is a backup sensor, the backup sensor is immediately triggered to re-collect the relevant data.

[0081] The second-level consistency check is: cross-validation of associated data. Its core rule is: when the difference between the reading of the contact core temperature sensor and the surface temperature inferred by the infrared thermal imager is continuously greater than expected (e.g., the difference is > ±2℃ in two consecutive samplings), it is determined to be a sensor deviation, and the system automatically switches to the backup sensor to re-collect relevant data and records the deviation value.

[0082] The third-level stability verification is as follows: set a fluctuation threshold (e.g., ±10%) for fluctuating data (e.g., cooling medium flow rate). If the fluctuation exceeds the threshold multiple times (e.g., 3 times) consecutively, it is judged as flow field fluctuation, the fluctuating data is marked and a fluctuation label is added.

[0083] Remove invalid data that has passed verification, and retain valid data and tagged fluctuation data, in the format of "timestamp (accurate to second) + data type (cooling medium / environment / product) + parameter name + value + unit" (e.g., "2020-05-20 18:23:47_cooling medium_flow rate_1.2_m"). 3 / h”), which are integrated into a structured and traceable multi-source fusion sensing set.

[0084] Methods for determining the current cooling phase using multi-condition logic include:

[0085] From the multi-source fusion perception, three core feature parameters were selected and extracted, including the product's average core temperature, cooling time, and real-time cooling rate.

[0086] Specifically, the arithmetic mean of all product core temperatures collected by temperature sensors is taken as the product average core temperature; the system time from the moment the product enters the cooling zone until the current moment is taken as the real-time cooling duration; the core temperature at the moment the product enters the cooling zone is taken as the initial core temperature, and the real-time cooling rate is equal to the ratio of the difference between the initial core temperature and the current core temperature to the cooling duration.

[0087] The extracted feature parameters are denoised (using a triple smoothing method to remove instantaneous jump values) to ensure the stability and accuracy of the feature parameters.

[0088] The cooling stage is divided into three categories: pretreatment stage, rapid cooling stage, and temperature equalization stage.

[0089] The multi-condition logic is defined as follows: based on the preset stage threshold range of the feature parameters, the threshold of the feature parameters is determined by the "AND" logic of the logic gates "AND", "OR", and "NOT" to identify the current cooling stage;

[0090] Specifically, the preset stage threshold range is determined by dividing the three feature parameters into three intervals based on the maximum and minimum thresholds of the corresponding feature parameters, corresponding to three cooling stages.

[0091] Multi-condition logic judgment:

[0092] The system is considered to be in the preprocessing stage if and only if all characteristic parameters simultaneously satisfy the following: cooling duration ≤ minimum cooling duration threshold (e.g., 5 minutes), average core temperature ≥ maximum core temperature threshold (e.g., 70℃), and real-time cooling rate ≥ maximum cooling rate threshold (e.g., 3.5℃ / min).

[0093] A rapid cooling phase is determined to be in effect if and only if all characteristic parameters simultaneously satisfy the following: cooling duration > minimum cooling duration threshold and ≤ maximum cooling duration threshold (e.g., 5-20 minutes); average core temperature > minimum core temperature threshold and < maximum core temperature threshold (e.g., 30-70℃); and real-time cooling rate > minimum cooling rate threshold and << maximum cooling rate threshold (e.g., 2.5-3.5℃ / min).

[0094] The product is considered to be in the uniform temperature stage if and only if all characteristic parameters simultaneously satisfy the following conditions: cooling duration > maximum cooling duration threshold (e.g., 20 minutes), average core temperature ≤ minimum core temperature threshold (e.g., 30℃), and real-time cooling rate ≤ minimum cooling rate threshold (e.g., 1.5℃ / min).

[0095] If parameters overlap during the threshold determination process (e.g., the cooling time meets the requirements for rapid cooling, but the average core temperature of the product is below 30°C), the average core temperature of the product shall be the primary criterion for determination.

[0096] The current cooling stage is determined by multi-condition logic. The determined stage type and corresponding feature parameters are packaged to generate a stage determination result package, which serves as the direct basis for subsequent calls to the gradient temperature control curve.

[0097] Methods for feedforward pre-compensation of environmental and product initial state include:

[0098] Based on the stage determination result package, according to the current cooling stage type, the reference gradient temperature control curve for the corresponding stage is called from the preset stage-gradient curve mapping library.

[0099] Among them, in the stage-gradient curve mapping library, the reference gradient temperature control curve for each type of cooling stage is preset according to the temperature change trend from initial target temperature to final target temperature, the reference cooling rate and reference duration matching the stage characteristics, so that the appropriate general reference gradient temperature control curve can be quickly located and called according to the stage type.

[0100] Example of a baseline gradient temperature control curve from the stage-gradient curve mapping library:

[0101] Pretreatment stage: The baseline gradient temperature control curve is from initial temperature 90℃ to target temperature 70℃, with a baseline cooling rate of 5℃ / min and a curve duration of 5 minutes.

[0102] Rapid cooling phase: The baseline gradient temperature control curve is from initial temperature 70℃ to target temperature 30℃, with a baseline cooling rate of 2.7℃ / min and a curve duration of 15 minutes.

[0103] Temperature equalization stage: The baseline gradient temperature control curve is from initial temperature 30℃ to target temperature 8℃, with a baseline cooling rate of 1.5℃ / min. The curve duration is dynamically adjusted according to the core temperature of the product (usually 10-15 minutes).

[0104] Environmental parameters and initial product parameters are extracted from multi-source fusion sensing.

[0105] Environmental parameters include the real-time ambient temperature and relative humidity of the cooling area; initial product parameters include the average unit weight of the product (the average unit weight of all products in the same batch) and the temperature at which the product exits the fryer.

[0106] The validity of the extracted data is verified a second time. Specifically, the marked fluctuation data is removed. For the empty data positions after the data removal, the average value of the parameters in the same batch is used to replace them, so as to ensure the reliability of the basic data for compensation calculation.

[0107] Feedforward pre-compensation is defined as: correcting the reference gradient temperature control curve according to preset compensation rules based on environmental parameters and initial product parameters;

[0108] Specifically, the logical definition of the compensation rules includes environmental compensation and product initial state compensation. The two types of compensation are calculated independently but take effect together.

[0109] Environmental compensation includes environmental temperature compensation and environmental humidity compensation; initial product state compensation includes product attribute compensation and previous process compensation.

[0110] Ambient temperature compensation is defined as follows: if the real-time ambient temperature deviates from the preset standard ambient temperature, the correction value of the baseline cooling rate is calculated according to the deviation range and the preset compensation coefficient (usually set to 0.1, but can be set according to the actual situation). The product of the compensation coefficient and the deviation range is taken as the correction value. The larger the deviation range, the correction value is adjusted synchronously according to the coefficient.

[0111] For example, ambient temperature compensation: Assuming the temperature compensation coefficient is set to 0.1 and the standard ambient temperature is 25℃, if the ambient temperature is >25℃, the baseline cooling rate increases by 0.1℃ / min for every 1℃ increase; if the ambient temperature is <25℃, the baseline cooling rate decreases by 0.1℃ / min for every 1℃ decrease.

[0112] The environmental humidity compensation is defined as follows: if the real-time relative humidity of the environment deviates from the preset standard environmental humidity, the correction value of the trigger threshold of the spray moisturizing system is adjusted according to the deviation range (when the deviation range meets the preset conditions, the trigger threshold is adjusted proportionally).

[0113] For example, environmental humidity compensation: Assuming the standard environmental humidity is 60%, an adjustment is triggered for every 5% decrease in the deviation, with an adjustment ratio of 2%. If the relative humidity is <60%, the trigger threshold (epidermal moisture content) of the spray moisturizing system is lowered by 2% for every 5% decrease (e.g., the standard threshold 55% → 53%). If the humidity is >60%, the threshold remains unchanged.

[0114] Product attribute compensation is defined as follows: if the average unit weight of the product deviates from the preset standard unit weight of the product, the correction value of the benchmark cooling rate is calculated according to the deviation range and the preset compensation coefficient (the product of the compensation coefficient and the deviation range is taken as the correction value).

[0115] For example, product attribute compensation: Assuming that an adjustment is triggered for every 5g increase or decrease in deviation, the attribute compensation coefficient is 0.01 when increasing and 0.008 when decreasing. If the average unit weight of the product is >50g (standard unit weight of the product), the baseline cooling rate increases by 0.05℃ / min for every 5g increase; if the unit weight is <50g, the baseline cooling rate decreases by 0.04℃ / min for every 5g decrease.

[0116] The compensation for the preceding process is defined as follows: if the temperature of the frying process deviates from the preset reference parameters of the preceding process, the initial temperature of the reference gradient temperature control curve is adjusted according to the deviation range (when the deviation range meets the preset conditions, the initial temperature node is finely adjusted according to the preset rules).

[0117] For example, the temperature compensation for frying: Assuming the temperature reduction ratio is 5℃ and the temperature increase ratio is 3℃, if the frying temperature is >95℃ (standard frying temperature), the initial temperature is reduced by 5℃ (e.g., the initial temperature in the rapid cooling stage is reduced from 70℃ to 65℃); if the frying temperature is <85℃ (the preset minimum frying temperature), the initial temperature is increased by 3℃.

[0118] The correction values ​​of environmental compensation and product initial state compensation are integrated to generate a compensated gradient temperature control curve (including the corrected temperature, target cooling rate, and curve duration). The calculation coefficients of each compensation item (such as the environmental temperature compensation coefficient of 0.1℃ / min·℃) and compensation rules are recorded simultaneously and packaged into a gradient curve-compensation result package, which serves as the core basis for the execution of the dynamic priority strategy.

[0119] The methods for generating gradient temperature control decision instruction sets include:

[0120] Based on the gradient curve-compensation result package and batch risk identifier, safety, quality, and efficiency are preset as control objectives. First, basic priority weights are assigned to the three control objectives according to the cooling stage (the total weight is 100%, used to adapt to the core needs of each stage). The specific basic priority weight allocation rules are as follows:

[0121] The pre-processing stage is defined as follows: with the core objective of quickly entering a stable cooling state, the basic priority weights are allocated in the order of efficiency > safety > quality.

[0122] The rapid cooling phase is defined as follows: with the core focus on controlling the duration of the critical risk range (the duration of the product's residence at 50-80℃), the basic priority weights are allocated according to safety > quality > efficiency.

[0123] It should be noted that research shows that simple HAs (heterocyclic amines, a type of heat-induced harmful substance) already formed in fried chicken feet are prone to structural transformation in the 50-80℃ range, forming more toxic derivatives. The transformation rate peaks at 65℃. In traditional cooling, it takes 1.5-2 hours for chicken feet to cool from 80℃ to 50℃. This prolonged residence time in this range leads to a significant increase in the total amount of HAs compared to the end of frying, and the proportion of derivatives also doubles. Therefore, it is necessary to control the duration of the product temperature within the critical risk range during the rapid cooling stage to reduce safety hazards.

[0124] The temperature equalization stage is defined as follows: with the core objective of ensuring the consistency of the final product state, the basic priority weights are allocated according to quality > safety > efficiency.

[0125] Then, the basic priority weights are fine-tuned based on batch risk identification to ensure that the control objectives for high-risk batches are strengthened. Specifically:

[0126] If a batch is identified as high-risk: Based on the basic priority weight of the corresponding cooling stage, increase the weight of the safety control target (the increase is preset based on the actual situation through expert experience, such as increasing proportionally (e.g., 5%, 10%) or increasing by a fixed value (e.g., 0.05%, 0.1), which is positively correlated with the risk level), and simultaneously reduce the weight of the efficiency control target.

[0127] If the batch risk is marked as normal: keep the basic priority weights of each cooling stage unchanged, and only slightly optimize the weights of the quality control targets based on the temperature deviation in the parameters of the compensated gradient temperature control curve (e.g., if the temperature fluctuation is small, fine-tune the weights to a balanced state).

[0128] Real-time verification of whether there is a conflict between the control objectives corresponding to the priority weights (i.e., the temperature control requirements corresponding to different control objectives are contradictory). If there is a conflict, the conflict between the control objectives is handled through the set general arbitration rules.

[0129] Specifically, the logic of the general arbitration rules is as follows:

[0130] If safety control objectives conflict with quality control objectives (e.g., increasing the cooling rate based on safety requirements (such as shortening the product's dwell time in critical risk areas), but decreasing the cooling rate based on quality requirements (such as avoiding product texture damage): prioritize safety control objectives and adopt the parameter requirements corresponding to safety control objectives (e.g., increasing the cooling rate based on safety requirements). Simultaneously, trigger quality protection auxiliary logic (i.e., synchronously adjust the parameters of auxiliary actuators such as spray moisturizing and adjustable flow guide, and adjust the parameters according to preset logic (e.g., increase the spray frequency to alleviate the dry and hardened skin caused by excessively rapid cooling, and adjust the angle of the flow guide to optimize temperature uniformity) to control quality damage within the preset acceptable range and mitigate the impact on quality).

[0131] If efficiency control objectives conflict with safety / quality control objectives (e.g., efficiency requirements (e.g., shortening the total cooling time) necessitate compressing the cooling stage curve duration, but safety control objectives (e.g., the dwell time in the critical risk zone is not less than a minimum) or quality control objectives (e.g., the duration of the temperature equalization stage is not less than a minimum) require extending the duration: prioritize meeting the safety / quality control objectives, adopting the parameter requirements corresponding to the safety / quality control objectives (e.g., extending the dwell time in the risk zone according to safety requirements), and abandoning the time reduction requirement of efficiency control objectives. If the efficiency loss exceeds the preset acceptable range (e.g., the total time extension exceeds 10%), trigger subsequent process coordination adjustments (send a cooling time extension warning to the downstream process through the MES system to reserve time for the downstream process to adjust its pace) to avoid affecting overall production efficiency;

[0132] After the arbitration is completed, an arbitration result package is generated, which contains three types of core information: the final value of the adjusted control target parameters (such as the determined cooling rate and the duration of each stage, which has been adapted to high-priority targets); the triggered auxiliary actuator action commands (such as the start-stop logic of the spray system and the adjustment rules of the deflector); and the arbitration record (conflict scenario type, arbitration rules, and parameter adjustment range), which is used for reference in subsequent iterative optimization.

[0133] The arbitrated control target parameters are combined with the actuator characteristics and the parameters of the compensated gradient temperature control curve to transform them into general control commands for each actuator.

[0134] Specifically, each actuator is categorized by function (refrigeration, humidification, auxiliary cooling, and media regulation), and the target parameters are converted into specific control commands for each actuator. The commands are expressed in a general form of action logic + trigger conditions (since the numerical logic differs among actuators, the conversion here does not involve specific numerical values):

[0135] For refrigeration actuators (such as variable frequency chillers): the command is converted into a target operating state command (including power regulation logic matched with the cooling rate). The command generation logic is to determine the power regulation logic based on the target cooling rate and the actuator characteristic library. For example, when the actual cooling rate is less than the target value, the power is increased by a preset step size; when the actual rate is greater than or equal to the target value, the current power is maintained, and a priority label is marked: safety / quality priority.

[0136] For moisturizing actuators (such as spray moisturizing systems): the instructions are converted into trigger conditions and operating mode commands (including start-stop logic that matches the quality control objectives). The command generation logic is to determine the start-stop trigger conditions and operating mode based on the quality protection requirements and epidermal state parameters. For example, when the epidermal moisture parameter is less than the target lower limit, the spray is started; when the target upper limit is reached, the spray is turned off. The modes are divided into intermittent spray and continuous spray (selected according to the quality protection requirements in the arbitration results).

[0137] For auxiliary cooling actuators (such as liquid nitrogen auxiliary modules): the instructions are converted into activation threshold and adjustment logic commands (including triggering rules that match the safety control objectives). The command generation logic is to determine the activation threshold and adjustment rules based on the safety objective requirements (such as risk zone dwell time warning). For example, when the risk zone dwell time is greater than the warning value, liquid nitrogen is activated; when the dwell time is less than or equal to the safety value, liquid nitrogen is deactivated. The adjustment rule is to adjust the liquid nitrogen flow rate proportionally based on the difference between the actual dwell time and the safety value.

[0138] For media control actuators (such as adjustable guide vanes): the state adjustment logic is converted into a state adjustment logic instruction (including adjustment rules that match temperature uniformity). The instruction generation logic is to determine the state adjustment logic based on the temperature uniformity requirement (the core sub-requirement of the quality target)—for example, when the temperature difference of the product surface temperature field is greater than the preset value, the guide vane is adjusted according to the preset angle step size; when the temperature difference is less than or equal to the preset value, the current angle of the guide vane is maintained.

[0139] The control commands of all actuators are structurally integrated to generate a gradient temperature control decision command set;

[0140] The instruction structure is defined as: unique executor identifier - control logic - trigger condition - priority identifier - associated target;

[0141] Example format: Variable frequency chiller unit - power adjustment (increase when the rate is low, maintain when the target is met) - based on comparison of actual cooling rate and target value - safety priority - associated with safety targets.

[0142] The ways to drive the actions of each actuator include:

[0143] Based on the temperature control decision instruction set, the actuator type and corresponding control logic of each actuator are extracted; the priority identifier (safety priority, quality priority, efficiency priority) and triggering conditions of each instruction are identified;

[0144] Establish a mapping relationship between instructions and associated control objectives (such as a certain instruction corresponding to the risk zone control in the safety control objective) to ensure that the execution logic is completely consistent with the control objective;

[0145] Then, instructions are scheduled to be executed in the order of priority identifier → actuator type, thereby strictly ensuring that high-priority actions are responded to first;

[0146] Priority sorting: First execute the instructions with safety priority (such as the liquid nitrogen auxiliary module activation instruction), then execute the instructions with quality priority (such as the start / stop instruction of the spray humidification system), and finally execute the instructions with efficiency priority (such as the power fine-tuning instruction of the refrigeration unit).

[0147] Refrigeration actuators: Based on the corresponding power regulation rules, the control module outputs corresponding regulation signals (such as frequency conversion signals) to drive them to adjust their operating status according to logic;

[0148] Humidification actuators: Based on the corresponding start / stop trigger conditions, they output switch signals to control the actuator to start or stop when the conditions are met;

[0149] Auxiliary cooling actuators: trigger auxiliary cooling functions (such as flow regulation of liquid nitrogen modules) according to the corresponding activation threshold and adjustment rules;

[0150] Media control actuators: Drive the regulating mechanism (such as the angle adjustment component of the guide vane) to perform corresponding actions according to the corresponding state adjustment logic;

[0151] During the actuator's operation, real-time operating status data is collected through sensors or the actuator's built-in feedback interface. The data collection frequency matches the action response frequency, including: the actuator's actual operating parameters (such as the current power of the refrigeration unit, the current operating mode of the spray system, the real-time value of liquid nitrogen flow, and the current angle of the guide vane), action response status (such as the actuator's response time after the command is issued, and whether the action is successfully executed), and abnormal status indicators (such as actuator action timeout, parameters exceeding the safe range, marked as abnormal and the cause of the abnormality recorded in detail).

[0152] The methods for generating gradient temperature control execution results include:

[0153] Collect core product data (such as core temperature and internal texture parameters) at a preset high-frequency sampling frequency (to adapt to the characteristics of the cooling stage, such as a short sampling interval in the rapid cooling stage and a slightly longer interval in the temperature equalization stage) to ensure that dynamic temperature changes can be captured.

[0154] By using specialized detection components (such as near-infrared and image recognition modules) to collect product surface data (such as surface moisture and texture integrity parameters), it is possible to avoid surface damage going undetected in a timely manner.

[0155] The collected data is aligned with the actual operating status data by timestamp, and the corresponding cooling stage is marked to form a set of feedback data containing the relationship between actuator action and product status.

[0156] Then, the real-time parameters in the feedback data set are compared with the target parameters (i.e., the changing trend of the target core temperature, the target cooling rate, etc.) in the compensated gradient temperature control curve to determine the parameter deviation. The deviation determination logic is linked with the compensation rules. Specifically:

[0157] The first step is to set an allowable deviation range for each parameter (such as cooling rate, core temperature, and skin moisture) (i.e., a general range that adapts to product characteristics, such as cooling rate deviation and core temperature deviation not exceeding the corresponding allowable deviation range).

[0158] The second step is to calculate the difference between the real-time parameters and the target parameters. If the difference is within the allowable deviation range, it is determined that there is no deviation and the current actuator action is maintained. If the difference exceeds the range, it is determined to be a deviation, and the deviation type (including cooling rate deviation (such as cooling rate being too slow) and product surface condition deviation (such as surface moisture being too low)) and deviation magnitude are recorded.

[0159] Based on the determined deviation type, targeted compensation instructions are generated according to the categorized compensation strategy to drive the actuator to adjust its actions. Simultaneously, it is ensured that the compensation does not violate the control target priority, and the compensation action is precisely matched with the deviation type. Specifically:

[0160] If the deviation is a cooling rate deviation: prioritize adjusting the refrigeration actuators, and fine-tune the refrigeration power according to the deviation magnitude by a preset step size (e.g., increase the power if the rate is low, and decrease the power if the rate is high); if the deviation is still not eliminated after refrigeration adjustment, and the current stage allows the use of auxiliary cooling (e.g., rapid cooling stage), then adjust the auxiliary cooling actuators (e.g., increase the liquid nitrogen flow rate, shorten the auxiliary cooling trigger delay) until the rate returns to the allowable deviation range.

[0161] If the deviation is due to the product's surface condition: prioritize adjusting the moisturizing actuators, increasing the spray frequency or extending the duration of each spray according to the deviation, and simultaneously adjusting the medium flow field through the guide vane actuator to avoid excessive local moisture loss; if the surface deviation is related to the cooling rate (e.g., excessively rapid cooling leading to moisture loss), without exceeding safety targets, slightly adjust the cooling actuator parameters (e.g., appropriately reduce the cooling rate) to balance the surface condition and cooling efficiency.

[0162] The compensation constraints are defined as follows: all compensation actions must be marked with compensation priority to ensure consistency with the priority of the control target (e.g., safety-related compensation takes precedence over quality-related compensation), and the compensation range shall not exceed the preset safety limit (e.g., the adjustment of cooling power shall not exceed a certain percentage of the rated power) to avoid over-compensation leading to new risks.

[0163] After the compensation command is executed, shorten the sampling interval (e.g., halve the original interval), collect the core data and surface data of the product again, verify the compensation effect and handle anomalies: if the real-time parameters return to the allowable deviation range, the compensation is deemed effective, the compensation command (e.g., actuator adjustment range, adjustment duration) is recorded, and the normal sampling frequency is restored; if the parameters still exceed the range after multiple consecutive compensations, the compensation is deemed ineffective, and anomaly handling is triggered. First, mark the compensation anomaly status, then simultaneously reduce the priority weight of non-critical control targets (e.g., efficiency targets), and prioritize safety and quality (e.g., extend the current stage duration, activate the backup cooling path) to prevent further deviation.

[0164] The execution process data (actual cooling curves at each stage, actuator action records (including initial actions and compensation adjustments), deviation types and compensation measures, current product core temperature, current surface moisture, and dwell time in critical risk zones) and stage status data (whether the current cooling stage is completed, whether the product status meets the target (e.g., whether the core temperature is close to the stage's final temperature), and whether there are any unresolved anomalies) are integrated to form the gradient temperature control execution results.

[0165] Methods for creating collaborative data and batch temperature control records and archiving them to a historical database include:

[0166] Key judgment parameters are extracted from the gradient temperature control execution results, including core product status parameters (i.e., core temperature), dangerous zone control parameters (i.e., dwell time in critical risk zones), product skin status parameters (i.e., skin moisture), and stage status data.

[0167] The judgment is executed according to the preset cooling completion standard. The judgment logic needs to cover both safety and quality dimensions. Specifically: if all parameters meet the preset standard at the same time (such as the core temperature meets the cooling final temperature requirement and the continuous sampling is stable, the dwell time in the risk range is ≤ the preset safety upper limit, the skin state parameters are within the qualified range, and each stage is completed), the cooling is judged to be complete; if any parameter does not meet the standard (such as the core temperature not reaching the final temperature, or the skin moisture is too low), the cooling is judged to be incomplete, and secondary compensation is triggered (based on the current deviation, a supplementary compensation instruction is generated to drive the actuator to continue to adjust until the standard is met);

[0168] If the process is deemed complete, the cooling completion process is triggered, and a cooling process report is generated, which includes process traceability data and core result parameters.

[0169] The process traceability data includes the actual temperature control curves for each cooling stage, all actuator action records (including initial actions and all compensation adjustments), deviation types and details of compensation measures (such as cooling rate deviation and corresponding cooling power adjustment range).

[0170] The core parameters of the results include the final cooling parameters (including the final core temperature and the final skin moisture), the risk zone control results (i.e., the comparison between the actual residence time in the risk zone and the preset upper limit), and the product status qualification mark (i.e., whether it meets the basic requirements of the product after cooling).

[0171] Send the cooling process report to the control terminal of the subsequent braising process, along with process adaptation suggestions (e.g., recommend the initial braising temperature based on the final core temperature, and recommend the braising humidity based on the final skin moisture). Then, receive parameter adaptation feedback from the subsequent braising process (e.g., whether the recommended process adaptation suggestions are accepted, and whether the final core temperature of the cooling process needs to be adjusted to match the braising rhythm). The feedback data must be associated with the batch unique identifier.

[0172] Integrate cooling process reports, parameter adaptation feedback, multi-source fusion sensing sets, gradient temperature control decision command sets, and all actuator action records, and link them together according to the unique identifier of the same batch to form complete cross-process collaborative data;

[0173] The integrated collaborative data is categorized into process archives, result archives, and collaborative archives to generate batch temperature control archives, which are then archived to the historical database. At the same time, an archive index is generated to facilitate quick access to data from multiple batches.

[0174] The process archive contains execution information for each cooling stage, the result archive contains cooling final state parameters and qualification indicators, and the collaboration archive contains cross-process adaptation suggestions and parameter adaptation feedback.

[0175] The formation methods of the entire process chain include:

[0176] Extract full-chain data from multiple batches (such as the most recent 30 batches) from historical databases, including three types of data: basic data, process data, and results;

[0177] The basic data includes a multi-source fusion sensing set and a gradient curve-compensation result package; the process data includes gradient temperature control execution results; and the result data includes cooling process reports and cross-process feedback.

[0178] The extracted end-to-end data is preprocessed, including cleaning, normalization, and classification labeling;

[0179] Among these steps, cleaning involves removing invalid data (such as outliers caused by sensor malfunctions or incomplete data from batch interruptions) and filling in occasional missing values ​​with the average of batches of the same type; normalization involves normalizing data of different dimensions (such as cooling rate, energy consumption, and cooling time) according to a preset standard range (such as a 0-100 score) to eliminate the impact of differences in units on the evaluation; and classification and labeling involves classifying data according to cooling stage and batch risk level (high risk / normal) to facilitate scenario-based evaluation and optimization.

[0180] Establish a four-dimensional evaluation mechanism encompassing safety, quality, efficiency, and energy consumption:

[0181] S91: Safety, quality, efficiency, and energy consumption are used as evaluation indicators, and the compliance rate for each indicator is defined. Specifically:

[0182] Safety dimension: The percentage of batches that meet safety standards (batches whose stay time in the risk zone is ≤ preset safety limit and have no safety warnings) out of the total number of batches is used as the safety compliance rate;

[0183] Quality dimension: The batches in which the core temperature and surface moisture of the product meet the standards after cooling are considered as qualified batches, and the proportion of these batches in the total batches is used as the quality compliance rate.

[0184] Efficiency Dimension: The proportion of batches whose actual cooling time is within the "target time ± preset deviation range" to the total number of batches is used as the efficiency compliance rate;

[0185] Energy consumption dimension: The proportion of batches with unit product cooling energy consumption (i.e., the ratio of total energy consumption to product output) lower than the industry benchmark value is used as the energy consumption compliance rate.

[0186] S92: Calculate the individual score for each evaluation indicator according to the mapping rule of compliance rate → score (that is, convert the compliance rate into a score, such as 100% safety compliance rate gets 100 points, 90% gets 90 points), and obtain the batch comprehensive performance score by weighted summation. Select batches whose comprehensive scores do not meet expectations (that is, set a comprehensive performance score threshold, such as 80 points, batches whose comprehensive performance scores are lower than the threshold are not meeting expectations) as key optimization batches.

[0187] It should be noted that different batches have different production requirements (e.g., high-risk batches need to prioritize safety, while regular batches need to balance efficiency and quality). Therefore, pre-defined weighting rules are necessary to ensure that the evaluation aligns with actual needs. For example, the weighting logic for high-risk batches is to balance quality and efficiency, while also considering safety and energy consumption. Therefore, the weights for safety, quality, efficiency, and energy consumption can be set to 40%, 30%, 15%, and 15%, respectively. Conversely, the weighting logic for regular batches prioritizes safety, followed by quality, with less emphasis on efficiency and energy consumption. Therefore, the weights can be set to 20%, 35%, 30%, and 15%, respectively.

[0188] S93: Conduct root cause analysis on key optimization batches to pinpoint performance shortcomings;

[0189] Specifically, the core objective of the assessment is to identify problems and then determine the direction, which requires breaking down and analyzing low-scoring batches or low-scoring dimensions.

[0190] Based on the actual scenario, root cause analysis rules are set through expert experience. For example, if the efficiency dimension score of the rapid cooling phase is only 60 points, it indicates that there is a problem with the efficiency control in this phase; if the safety dimension score of the high-risk batch is 85 points (lower than the average score of 95 points for regular batches), it indicates that the safety strategy for the high-risk batch needs to be strengthened.

[0191] Then, based on the shortcomings, historical data was used to trace back to specific control links. For example, if the efficiency score was low, it was further broken down into the fact that the actual duration of the rapid cooling phase exceeded the target by 20% → further traced back to insufficient feedforward pre-compensation (the target rate was not corrected in time when the ambient temperature rose). The core shortcoming was identified as "insufficient adaptability of feedforward pre-compensation parameters".

[0192] For example, a low score in the quality dimension can be further broken down into dry and hard skin leading to subsequent complaints → traced back to insufficient frequency of moisturizing compensation (the spray was not triggered in time due to the deviation in skin moisture). The core shortcoming is that the threshold parameter for moisturizing compensation moisture deviation is not accurate enough.

[0193] To address performance shortcomings, an optimization scheme was developed based on simulation verification, including gradient temperature control curves, feedforward pre-compensation parameters, and priority weights. Specifically:

[0194] The gradient temperature control curve is optimized as follows: if the deviation rate between the actual temperature curve and the target curve in a certain stage is greater than the preset value, the correlation between ambient temperature and curve deviation and product weight and curve deviation is analyzed by linear regression. The temperature node interval of the curve and the segmentation coefficient of the target rate are adjusted (e.g., in the rapid cooling stage, the single rate is adjusted to a high rate in the early stage and a low rate in the later stage, so as to balance efficiency and quality).

[0195] The feedforward pre-compensation parameters are optimized as follows: if high-frequency deviations still occur after feedforward compensation, the compensation coefficient is corrected (e.g., the rate increases by 0.1℃ / min for every 1℃ increase in ambient temperature is adjusted to 0.12℃ / min for every 1℃ increase), or a new compensation dimension is added (e.g., a compensation item for the cooling rate based on the initial surface moisture of the product is added).

[0196] The priority weighting has been optimized as follows: if the safety target of a high-risk batch is not met, the weighting of the safety dimension of the high-risk batch will be increased (e.g., from 35% to 45%). At the same time, the conflict arbitration rules will be optimized (e.g., when safety and quality conflict, safety will be given priority while adding epidermal moisturizing auxiliary measures).

[0197] Using simulation verification tools, a simulation verification model is built based on typical batch data from historical databases (such as three representative batches each of high-risk and regular batches) to simulate the execution effect of optimized parameters.

[0198] If the simulation results meet the requirement of an overall performance score improvement of ≥10% and no decrease in the compliance rate of each dimension, the optimization scheme is deemed effective; if the requirement is not met, return to adjust the optimization parameters (such as fine-tuning the compensation coefficient and correcting the curve nodes) until the simulation passes.

[0199] Based on the optimized scheme verified by simulation, the gradient temperature control decision instruction set is updated and used as the initial judgment basis for the next round of multi-condition logic (i.e., adjusting the corresponding feature parameter thresholds, such as adjusting the core temperature judgment threshold and cooling rate judgment range in the preprocessing stage, so as to ensure that the entire link from data acquisition to stage judgment in the next round is adapted to the optimization strategy), forming a data-driven-evaluation optimization-closed-loop iteration full-process link.

[0200] Example 2

[0201] Please see Figure 3 As shown, the parts not described in detail in this embodiment are described in Embodiment 1. A gradient temperature control method for automated processing of tiger-skin chicken feet is provided, including:

[0202] S1: Real-time collection of data related to cooling medium, environment and products, combined with data from the previous frying process, and generated a multi-source fusion sensing set through three-level verification;

[0203] S2: Based on a multi-source fusion sensing set, the current cooling stage is determined through multi-condition logic, the corresponding gradient temperature control curve is called, and feedforward pre-compensation is performed for the initial state of the environment and product. Combined with batch risk identification, a dynamic priority strategy is executed to generate a gradient temperature control decision instruction set.

[0204] S3: Based on the gradient temperature control decision instruction set, drive the actions of each actuator, collect the core temperature and surface data of the product in real time for feedback compensation, and generate gradient temperature control execution results.

[0205] S4: Based on the results of gradient temperature control, when the core temperature of the product meets the expectation, cooling is triggered to complete, a cooling report is generated and sent to the subsequent brining process and feedback is received, forming collaborative data and batch temperature control files and archiving them to the historical database;

[0206] S5: Based on multiple batches of data from the historical database, through multi-dimensional performance evaluation, optimize the gradient temperature control curve and feedforward pre-compensation parameters, update the gradient temperature control decision instruction set, and use it as the initial judgment basis for the next round of multi-condition logic, forming a full-process optimization link.

[0207] Example 3

[0208] This embodiment discloses an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the operation mode of the aforementioned automated processing gradient temperature control system for tiger skin chicken feet.

[0209] Since the electronic device described in this embodiment is the electronic device used to implement the gradient temperature control method for automated processing of tiger-skin chicken feet in this application embodiment, those skilled in the art can understand the specific implementation method and various variations of the electronic device in this embodiment based on the gradient temperature control method for automated processing of tiger-skin chicken feet described in this application embodiment. Therefore, how the electronic device implements the method in this application embodiment will not be described in detail here. As long as those skilled in the art implement the electronic device used in the gradient temperature control method for automated processing of tiger-skin chicken feet in this application embodiment, it falls within the scope of protection of this application.

[0210] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters and thresholds in the formulas are set by those skilled in the art according to the actual situation.

[0211] The above description is merely a preferred embodiment of the present invention, and the scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for users of ordinary technical skills, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A gradient temperature control system for automated processing of tiger-skin chicken feet, characterized in that, include: Multi-source data acquisition and fusion module: Real-time acquisition of data related to cooling medium, environment and products, combined with data from the previous frying process, and through three-level verification, generates a multi-source fusion sensing set; Stage Judgment Decision Generation Module: Based on a multi-source fusion sensing set, the module determines the current cooling stage through multi-condition logic, calls the corresponding gradient temperature control curve, performs feedforward pre-compensation for the initial state of the environment and product, and combines batch risk identification to execute a dynamic priority strategy to generate a gradient temperature control decision instruction set. The generation methods of the gradient temperature control decision instruction set include: Based on gradient curve-compensation result package and batch risk identifier, safety, quality and efficiency are taken as control objectives. First, the basic priority weights of the three control objectives are allocated according to the cooling stage. Then, the basic priority weights are fine-tuned in a targeted manner in combination with batch risk identifiers to obtain the adjusted priority weights. Real-time verification of whether there are conflicts between the control objectives corresponding to the priority weights, and handling of existing control objective conflicts through the set general arbitration rules; The arbitrated control target parameters are combined with the actuator characteristics and the parameters of the compensated gradient temperature control curve to transform them into general control commands for each actuator. The common control instructions for all actuators are structured and integrated to generate a gradient temperature control decision instruction set; Actuator drive compensation module: Based on the gradient temperature control decision instruction set, it drives the actions of each actuator, collects the core temperature and surface data of the product in real time for feedback compensation, and generates gradient temperature control execution results; The generation methods for the gradient temperature control execution results include: The core and surface data of the product are collected at high frequency, aligned with the actual operating status data by timestamp, and then compared with the compensated gradient temperature control curve to determine the parameter deviation. Based on the determined deviation type, a targeted compensation instruction is generated according to the categorized compensation strategy to drive the actuator to adjust its actions; After the compensation command is executed, the compensation effect is verified and anomalies are handled in a timely manner, generating gradient temperature control execution results that include execution process data and stage status data; Cooling completion determination module: Based on the results of gradient temperature control, when the core temperature of the product meets the expectation, cooling is triggered, a cooling report is generated and sent to the subsequent brining process and feedback is received, forming collaborative data and batch temperature control files and archiving them to the historical database; Iterative optimization module: Based on multiple batches of data from the historical database, it optimizes the gradient temperature control curve and feedforward pre-compensation parameters through multi-dimensional performance evaluation, updates the gradient temperature control decision instruction set, and serves as the initial judgment basis for the next round of multi-condition logic, forming a full-process optimization link.

2. The automated gradient temperature control system for processing tiger-skin chicken feet according to claim 1, characterized in that, The generation methods of the multi-source fusion sensing set include: Collect multi-source data from the cooling process and categorize it into cooling medium data, environmental data, and product data according to data type; simultaneously, use RFID readers to read batch risk identifiers and key parameters from the preceding frying process and merge them with product data to form product-related data; The three-level verification logic is defined as: Level 1 integrity verification, Level 2 consistency verification, and Level 3 stability verification. All acquired data undergoes three levels of verification to identify invalid data and mark fluctuating data, thereby retaining valid data and marked fluctuating data, and integrating them into a structured multi-source fusion sensing set.

3. The automated processing gradient temperature control system for tiger-skin chicken feet according to claim 2, characterized in that, The method of determining the current cooling stage through multi-condition logic includes: Based on a multi-source fusion sensing set, the product's average core temperature, cooling duration, and real-time cooling rate are extracted as core feature parameters. The cooling stage is divided into three categories: pretreatment stage, rapid cooling stage, and temperature equalization stage. The multi-condition logic is defined as follows: based on the preset stage threshold range of the feature parameters, the threshold of the feature parameters is determined through the AND logic of the logic gate to identify the current cooling stage; If parameters overlap during the threshold determination process, the average core temperature of the product shall be the primary criterion for determination. The current cooling stage is determined by multi-condition logic, and the determined stage type and corresponding feature parameters are packaged to generate a stage determination result package.

4. The automated gradient temperature control system for processing tiger-skin chicken feet according to claim 3, characterized in that, The methods for performing feedforward pre-compensation for the initial state of the environment and product include: Based on the stage determination result package, according to the current cooling stage type, the corresponding stage's benchmark gradient temperature control curve is called from the preset stage-gradient curve mapping library; at the same time, environmental parameters and product initial parameters are extracted from the multi-source fusion sensing set. Feedforward pre-compensation is defined as follows: by using environmental parameters and initial product parameters, and according to preset compensation rules, the reference gradient temperature control curve is corrected to generate a compensated gradient temperature control curve, and packaged into a gradient curve-compensation result package.

5. The automated gradient temperature control system for processing tiger-skin chicken feet according to claim 1, characterized in that, The methods for driving the actuators include: Based on the temperature control decision instruction set, the actuator type, control logic, triggering conditions, and control target priority of each instruction are extracted for each actuator. Then, various actuators are driven to perform corresponding actions according to the priority order of the control target, while real-time data of the actuator's operating status is collected and recorded.

6. The automated gradient temperature control system for processing tiger-skin chicken feet according to claim 1, characterized in that, The methods for generating collaborative data and batch temperature control records and archiving them to the historical database include: Based on the results of gradient temperature control, key judgment parameters are extracted, and judgment is performed according to the preset cooling completion standard. If the judgment is not completed, secondary compensation is triggered until the cooling is completed. If the process is deemed complete, the cooling completion process is triggered, a cooling process report containing process traceability data and core result parameters is generated, and the report is sent to the subsequent brining process and parameter adaptation feedback is received. Integrate cooling process reports and parameter adaptation feedback to form cross-process collaborative data and batch temperature control archives, which are then archived in a historical database.

7. The automated gradient temperature control system for processing tiger-skin chicken feet according to claim 6, characterized in that, The formation methods of the aforementioned end-to-end optimized link include: Extracting multiple batches of end-to-end data from historical databases to construct a four-dimensional evaluation mechanism: S71: Use safety, quality, efficiency and energy consumption as evaluation indicators, and define the compliance rate for each evaluation indicator. S72: Calculate the individual score for each evaluation indicator according to the mapping rule of compliance rate → score, and obtain the batch comprehensive performance score by weighted summation. Select batches whose comprehensive scores do not meet expectations as key optimization batches. S73: Conduct root cause analysis on key optimization batches to pinpoint performance shortcomings; To address performance shortcomings, optimization schemes for gradient temperature control curves, feedforward pre-compensation parameters, and priority weights were developed based on simulation verification. By updating the gradient temperature control decision instruction set through optimization and simultaneously using it as the initial judgment basis for the next round of multi-condition logic, a full-process optimization chain of data-driven, evaluation optimization, and closed-loop iteration is formed.

8. A gradient temperature control method for automated processing of tiger-skin chicken feet, implemented based on the gradient temperature control system for automated processing of tiger-skin chicken feet as described in any one of claims 1 to 7, characterized in that, include: S1: Real-time collection of data related to cooling medium, environment and products, combined with data from the previous frying process, and generated a multi-source fusion sensing set through three-level verification; S2: Based on a multi-source fusion sensing set, the current cooling stage is determined through multi-condition logic, the corresponding gradient temperature control curve is called, and feedforward pre-compensation is performed for the initial state of the environment and product. Combined with batch risk identification, a dynamic priority strategy is executed to generate a gradient temperature control decision instruction set. S3: Based on the gradient temperature control decision instruction set, drive the actions of each actuator, collect the core temperature and surface data of the product in real time for feedback compensation, and generate gradient temperature control execution results. S4: Based on the results of gradient temperature control, when the core temperature of the product meets the expectation, cooling is triggered to complete, a cooling report is generated and sent to the subsequent brining process and feedback is received, forming collaborative data and batch temperature control files and archiving them to the historical database; S5: Based on multiple batches of data from the historical database, through multi-dimensional performance evaluation, optimize the gradient temperature control curve and feedforward pre-compensation parameters, update the gradient temperature control decision instruction set, and use it as the initial judgment basis for the next round of multi-condition logic, forming a full-process optimization link.

Images