Energy-saving vegetable dehydration processing control method

Abstract

The application provides an energy-saving vegetable dehydration processing control method, and relates to the field of industrial automation control. The dehydration dynamics state matrix is constructed by synchronously acquiring the outlet humidity change information and the processing cavity pressure fluctuation information. The critical point of the constant-speed dehydration stage to the speed-reducing dehydration stage is identified according to the matrix, and the stage conversion trigger signal is generated. After responding to the trigger signal, the heating power and the fan speed are cooperatively and gradually decreased according to the attenuation degree of the current dehydration rate relative to the constant-speed stage, the dynamic energy input strategy is formed, and the execution mechanism is driven. The heating power and the fan speed are real-time matched with the current dehydration rate, the technical defects that the energy input and the actual dehydration rate are seriously mismatched due to the fact that the fixed time sequence switching strategy cannot accurately capture the real-time change point of the dehydration dynamics characteristics are overcome, and the energy utilization efficiency of the vegetable dehydration processing process is effectively improved.

Description

Technical Field

[0001] This invention relates to the field of industrial automation control, and more specifically, to an energy-saving vegetable dehydration processing control method. Background Technology

[0002] Vegetable dehydration is a crucial step in the intensive processing of agricultural products. Hot air drying dominates the large-scale dehydrated vegetable production in China due to its advantages such as low equipment cost, simple operation and maintenance, and large processing capacity. During hot air drying, the dehydration rate of vegetable materials typically exhibits distinct phases, generally categorized into an adjustment phase, a constant-rate dehydration phase, and a falling-rate dehydration phase. The constant-rate dehydration phase primarily removes surface free water and capillary water, resulting in a relatively high and stable dehydration rate. Upon entering the falling-rate dehydration phase, the migration of internal bound water becomes the dominant process, and the moisture diffusion rate decreases significantly. Based on this drying kinetic characteristic, the academic community has conducted extensive research on the construction of drying kinetic models, numerical simulation of heat and mass transfer, and optimization of different heating method combinations. However, existing research largely focuses on offline optimization of drying process parameters—determining optimal drying temperature, wind speed, and moisture content at stage transition points through orthogonal experiments or response surface methodology. Technical accumulation in online dynamic control of the drying process remains relatively weak.

[0003] Existing vegetable dehydration processing control systems generally employ fixed power or fixed timing operation modes. Specifically, the heating intensity and fan speed of the hot air circulation system typically operate according to a preset program, lacking an adaptive switching mechanism based on real-time status between the constant-speed dehydration stage and the dehydration stage. While some improved solutions recognize the differences in hot air parameter requirements at different drying stages and propose control approaches that set different temperatures or wind speeds in stages, their stage switching still relies on fixed time nodes or preset critical moisture content thresholds. This open-loop control strategy faces two inherent drawbacks: First, the initial moisture content, tissue density, and material thickness of different varieties and batches of vegetables vary significantly, resulting in large fluctuations in the critical point of transition from the constant-speed stage to the dehydration stage. Fixed timing or fixed threshold switching strategies cannot accurately capture the real-time transition points of dehydration kinetics. Second, after entering the dehydration stage, the migration rate of bound water within the material decreases significantly, and the actual dehydration rate has already decreased significantly. However, if the control system continues to maintain high-power heating and strong air circulation during the constant-speed stage, it will cause a severe mismatch between energy input and the actual dehydration rate, resulting in a large amount of ineffective energy consumption. The aforementioned problems directly hinder the improvement of energy efficiency in the dehydrated vegetable processing process. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides an energy-saving vegetable dehydration processing control method. This method can, to some extent, solve the problem that while the water migration rate is significantly reduced during the vegetable dehydration stage, existing control methods still maintain high-power heating and strong air circulation. Furthermore, the fixed timing switching strategy cannot accurately capture the transition point of the dehydration stage, resulting in a severe mismatch between energy input and dehydration rate, leading to a large amount of ineffective energy consumption.

[0005] According to one aspect of the present invention, an energy-saving vegetable dehydration processing control method is provided, comprising:

[0006] The humidity change information and pressure fluctuation information during the vegetable dehydration process are obtained, and a dehydration dynamic state matrix is ​​constructed based on the humidity change information and the pressure fluctuation information.

[0007] Based on the aforementioned dehydration kinetic state matrix, the critical point for the transition between dehydration stages is identified, and a stage transition trigger signal is generated;

[0008] In response to the stage transition trigger signal, the heating power and fan speed are coordinated and gradually reduced according to the degree of decay of the current dehydration rate relative to the dehydration rate of the constant rate dehydration stage, so as to obtain a dynamic energy input strategy.

[0009] The heating unit and fan actuator are driven according to the dynamic energy input strategy to match the heating power and fan speed with the current dehydration rate.

[0010] Furthermore, the humidity change information includes an air outlet humidity change sequence and a humidity change rate sequence obtained from the air outlet humidity change sequence, and the pressure fluctuation information includes a processing cavity pressure fluctuation sequence.

[0011] The humidity change rate sequence is obtained by pressure-weighted local regression differentiation. The pressure-weighted local regression differentiation includes: determining the degree of deviation between the pressure fluctuation corresponding to each humidity sampling point and the average pressure fluctuation in the local time window; determining the regression weight of each humidity sampling point according to the degree of deviation; and calculating the local regression slope of the air outlet humidity change sequence based on the regression weight.

[0012] Furthermore, the dehydration kinetic state matrix includes sampling time, outlet humidity content, humidity change rate, processing chamber pressure fluctuation, humidity change rate variance, pressure fluctuation variance, and data integrity marker; the continuous time window used to construct the dehydration kinetic state matrix is ​​jointly determined by the air turnover cycle of the hot air circulation system, sensor response time, and control cycle, and is rolled over and corrected based on the stage transition backtracking results after batch processing.

[0013] Furthermore, the critical point for the dehydration stage transition is determined by a stage transition index, which includes a humidity change rate decay term, a humidity change rate variance convergence term, a pressure fluctuation variance convergence term, and a trend persistence term.

[0014] The humidity change rate decay term is used to characterize the degree of decrease of the mean humidity change rate of the current window relative to the mean humidity change rate of the constant rate dehydration stable interval. The humidity change rate variance convergence term is used to characterize the degree of convergence of the variance of the humidity change rate of the current window relative to the variance of the humidity change rate of the constant rate dehydration stable interval. The pressure fluctuation variance convergence term is used to characterize the degree of convergence of the variance of the pressure fluctuation of the current window relative to the variance of the pressure fluctuation of the constant rate dehydration stable interval. The trend persistence term is used to characterize the persistence of the humidity change rate evolving in a downward direction within multiple consecutive windows.

[0015] Furthermore, the stage transition index corresponds to a stable sample band and a transition sample band;

[0016] The stable sample band includes the stage transition index statistical interval of the constant rate dehydration stable window confirmed by offline moisture content in historical qualified batches, and the turning sample band includes the stage transition index statistical interval of the stage transition window confirmed by offline moisture content in historical qualified batches.

[0017] When the stage transition index corresponding to the current window continuously enters the turning sample band and leaves the stable sample band, the stage transition trigger signal is generated; wherein, only batches with qualified final product moisture content and unit energy consumption are used to update the stable sample band and the turning sample band.

[0018] Furthermore, the dynamic energy input strategy includes residual energy demand factor, target heating power, and target fan speed;

[0019] The residual energy demand factor is obtained by coupling the ratio of the current humidity change rate to the mean humidity change rate during the constant-rate dehydration stage, the ratio of the current pressure fluctuation variance to the pressure fluctuation variance during the constant-rate dehydration stage, the stage transition index, and the cavity temperature deviation.

[0020] The target heating power is determined based on the residual energy demand factor and the minimum stable operating boundary of the heating unit, and the target fan speed is determined based on the residual energy demand factor and the minimum effective dehumidification operating boundary of the fan. The target heating power and the target fan speed have different decreasing response sensitivities.

[0021] Furthermore, the cavity temperature deviation is determined by the current cavity temperature and the target cavity temperature; the current cavity temperature includes the average temperature of the upper, middle and lower temperature measuring points inside the processing cavity;

[0022] When any temperature measuring point fails, the current cavity temperature is obtained by compensating for the average temperature of the remaining temperature measuring points with the historical offset of the failed temperature measuring point; the target cavity temperature is the constant process temperature corresponding to the current dehydration stage.

[0023] According to another aspect of the present invention, an energy-saving vegetable dehydration processing control system is provided, comprising:

[0024] The state perception and feature extraction module is used to acquire humidity change information and pressure fluctuation information during the vegetable dehydration process, and to construct a dehydration dynamic state matrix based on the humidity change information and the pressure fluctuation information.

[0025] The stage transition identification module is used to identify the critical point of dehydration stage transition based on the dehydration kinetic state matrix and generate a stage transition trigger signal.

[0026] The dynamic energy control module is used to respond to the stage transition trigger signal and, based on the degree of decay of the current dehydration rate relative to the dehydration rate of the constant-rate dehydration stage, to coordinately reduce the heating power and the fan speed to obtain a dynamic energy input strategy.

[0027] The execution and self-optimization module is used to drive the heating unit and the fan actuator according to the dynamic energy input strategy so that the heating power and fan speed are matched with the current dehydration rate.

[0028] Compared with existing technologies, the energy-saving vegetable dehydration processing control method provided by this invention simultaneously acquires information on humidity changes at the air outlet and pressure fluctuations in the processing chamber during the vegetable dehydration process. This constructs a dehydration dynamic state matrix that simultaneously reflects the strength of moisture migration and the state of airflow disturbance. Based on this state matrix, the critical point of transition from a constant-rate dehydration stage to a deceleration stage is identified, and a stage transition trigger signal is generated. Upon responding to the trigger signal, the heating power and fan speed are coordinated and gradually reduced according to the degree of attenuation of the current dehydration rate relative to the constant-rate stage, forming a dynamic energy input strategy and driving the actuator to ensure that the heating power and fan speed are matched with the current dehydration rate in real time. This overcomes the technical defect in existing technologies where the fixed-sequence switching strategy cannot accurately capture the real-time transition points of dehydration dynamic characteristics, resulting in a severe mismatch between energy input and the actual dehydration rate, effectively improving the energy utilization efficiency of the vegetable dehydration processing. Attached Figure Description

[0029] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings:

[0030] Figure 1 This is a flowchart of an energy-saving vegetable dehydration processing control method according to an embodiment of the present invention.

[0031] Figure 2 This is a schematic diagram illustrating the evolution of each component of the stage transition index in an energy-saving vegetable dehydration processing control method according to an embodiment of the present invention.

[0032] Figure 3 This is an application scenario diagram of the energy-saving vegetable dehydration processing control method according to an embodiment of the present invention. Detailed Implementation

[0033] Hereinafter, exemplary embodiments according to the present invention will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of the present invention, and not all embodiments of the present invention. It should be understood that the present invention is not limited to the exemplary embodiments described herein.

[0034] As mentioned in the background section, during the vegetable dehydration process, when vegetables enter the dehydration stage, a large amount of internal free water has evaporated, and the migration rate of the remaining bound water has decreased significantly. However, existing control methods still maintain high-power heating and strong air circulation during the constant-rate dehydration stage, resulting in a serious mismatch between the actual dehydration rate and energy input. At the same time, the critical moisture content of different batches of vegetables fluctuates when they transition from the constant-rate dehydration stage to the dehydration stage. The fixed-time stage switching strategy cannot accurately capture the real-time transition point of the dehydration kinetics, resulting in continuous excessive energy input without a synchronous increase in dehydration efficiency, leading to a large amount of ineffective energy consumption.

[0035] Figure 1 This is a flowchart of an energy-saving vegetable dehydration processing control method according to an embodiment of the present invention. Figure 1 As shown, the energy-saving vegetable dehydration processing control method includes:

[0036] S1. First, the original humidity change sequence and pressure fluctuation sequence within a continuous time window are simultaneously acquired from the air outlet humidity detection unit and the processing cavity pressure detection unit. When the humidity detection unit outputs relative humidity, it is first converted into absolute humidity or moisture content by combining the air outlet temperature and local air pressure. Then, the two data are zero-point corrected, time resampled, and timestamp aligned so that each sampling moment corresponds to a unique humidity value and pressure value. After alignment, a pressure weighting coefficient is introduced into the local causal window centered on the current sampling moment. Local regression differentiation is performed on the humidity sequence to obtain the humidity change rate sequence. The sampling moment, humidity value, humidity change rate, pressure value, window variance, and data integrity mark are written into the dehydration kinetic state matrix to form the online state basis required for subsequent stage identification.

[0037] Specifically, the local causal window is preferably one to three air turnover cycles. Initially, it can be determined based on the effective volume of the duct and the circulating air volume. Subsequent adjustments are made based on final product quality, stage switching backtracking results, and equipment response speed to ensure the window covers the actual moisture migration process without obscuring critical point changes due to excessive window length. The humidity change rate is not directly obtained from the difference between adjacent sampling points. Instead, pressure stability is used as a weight to correct the local regression. The closer the pressure fluctuation is to the local mean, the greater the contribution of that sample to the derivative estimation. The more pronounced the pressure abrupt change, the lower its weight, thus separating short-term anomalies caused by damper action, transient fan fluctuations, or local steam impacts from the actual moisture release trend. The pressure fluctuation sequence can use at least one of the following: differential pressure between the processing chamber and the environment, pressure difference between the fan inlet and outlet, or local pressure difference in the duct. This sequence is then corrected for no-load baseline before being used in the calculation to avoid equipment zero-point drift affecting state judgment. When the sampling periods of humidity and pressure sensors are inconsistent, the controller preferably uses the equipment control cycle as a unified time axis, performs window aggregation on high-frequency data, performs time interpolation on low-frequency data, and assigns lower confidence weights to interpolation points to ensure that subsequent local regression differentiation is not skewed by a single missing point. If a sampling point is affected by door opening, maintenance intervention, abnormal shutdown, or short-term communication interruption, the sampling point is retained in the state matrix but marked as a low-confidence sample, and its subsequent weight is automatically reduced to avoid its direct participation in stage identification.

[0038] In one embodiment, the humidity change rate can be determined according to the following formula:

[0039]

[0040]

[0041] in, Indicates the first The rate of change of humidity at each sampling time, in units of or kg water / Its physical meaning is the trend of change in humidity content at the air outlet per unit time; Represents the local fitting neighborhood Inner Humidity content at the air outlet of each sampling point; Indicates the sampling time, in minutes; and These represent the average values ​​of time and humidity content within the neighborhood, respectively. This indicates the pressure fluctuation at the corresponding sampling point, in Pa. and These represent the mean and standard deviation of the pressure fluctuation within this local neighborhood, respectively. It is a pressure sensitivity index used to adjust the pressure disturbance on the sample weight. Its value can be calibrated according to the vegetable type, material layer thickness, fan disturbance amplitude and historical stable batches. To avoid extremely small positive numbers with a denominator of zero, their dimensions are consistent with those of the denominator.

[0042] In the above formulas, the first equation generates weights, and the second equation is the weighted local regression slope. Together, they output a humidity change rate that more closely approximates the actual moisture release process. When constructing the dehydration kinetic state matrix, the controller stacks the sampling time, humidity content, humidity change rate, pressure fluctuation, humidity change rate variance, pressure fluctuation variance, and data integrity markers at each sampling moment in chronological order, forming a matrix that characterizes the evolution trajectory of the current batch's dehydration state. The humidity change rate variance describes the dispersion of moisture migration, the pressure fluctuation variance describes whether the cavity flow state converges, and the data integrity marker is used for natural weight reduction of abnormal samples during subsequent window analysis. After construction, the dehydration kinetic state matrix is ​​sent to the subsequent stage transformation identification process to simultaneously reflect the strength of moisture migration and the state of airflow disturbance on the same data basis. To ensure the subsequent parameters can be implemented, the pressure sensitivity index... The optimal method is to determine the calibration by comparing several batches of stable operating data and several batches of disturbed operating data, and selecting the value that results in the smoothest humidity change rate curve, the smallest stage switching lag, and qualified final product quality; for example, in leafy vegetables, The value can fall within the range of 1.0 to 1.5, and can be fine-tuned according to the specific air duct structure and material layer thickness. In this way, S1 not only completes data acquisition, but also maps the humidity migration trend and the degree of airflow disturbance into a state matrix that can be directly called by subsequent algorithms, avoiding the coarseness of traditional methods that rely on a single humidity value or a fixed time period to judge the stage.

[0043] S2. Subsequently, a sliding window variance analysis is performed on the dehydration kinetic state matrix to extract the mean humidity change rate, the variance of the humidity change rate, and the variance of the pressure fluctuation within each window. Combined with the online baseline formed by the current batch's constant-rate dehydration stable interval and the turning point sample band formed by historical qualified batches, a stage transition index is constructed to identify the critical point of transition from constant-rate dehydration to falling-rate dehydration. After confirmation, a stage transition trigger signal is output.

[0044] Specifically, the controller moves the analysis window along the time axis. The length of each analysis window is preferably determined by the air turnover time, sensor response time, and control cycle, ensuring that the window covers at least one complete hot air cycle. To balance response speed and stability, the windows can partially overlap to ensure that inflection points are continuously captured. Within each window, the controller calculates the mean humidity change rate, the variance of the humidity change rate, and the variance of pressure fluctuation, using the identified constant-rate dehydration stable interval for the current batch as the online baseline. The so-called constant-rate dehydration stable interval is not a fixed duration, but rather a set of continuous windows in the current batch where the mean and variance of the humidity change rate remain relatively stable, and the variance of pressure fluctuation does not show significant expansion. If a sufficiently stable set of windows has not yet been formed in the initial stage of the current batch, the controller first initializes it with the baseline of historical qualified batches of the same type and loading grade, and then gradually replaces it with the current batch baseline as online data accumulates. This avoids using fixed time series or fixed moisture content as switching conditions, instead allowing the inflection point judgment to be generated from the dynamic distribution of the current batch itself. For ease of engineering implementation, the stable sample band can be defined as... The transition sample band can be defined as ,in These results were obtained from statistical analysis of constant-rate stable windows with offline moisture content confirmation from historical qualified batches. These statistics are obtained from historical confirmed transition windows. When the system is running for the first time or the number of valid stable windows in the current batch is insufficient, the controller initializes the above statistics with historical samples. Once the number of stable windows in the current batch reaches a sufficient number, the statistics of the current batch are then used to adjust the data. Rolling updates are implemented to allow the sample bands to adaptively adjust to variations in raw materials and environmental changes. Only batches whose final product moisture content meets the standards and whose unit energy consumption is within acceptable limits are allowed to have their transition window statistics written into the historical sample database and used to update the transition sample bands. Unacceptable batches are marked as abnormal and removed, and are not included in the sample band updates to ensure the stability of the historical statistical basis. Here, "acceptable" is not an abstract term, but has a clear process meaning: the moisture content of the final product of this batch should fall within the process-allowed range, and the unit energy consumption should not be significantly higher than the median level of similar batches. If only the moisture content is acceptable but the energy consumption is abnormally high, this batch is still not suitable as a source for transition sample updates, as it may reflect equipment malfunctions, sensor drift, or uneven material layer distribution.

[0045] In one implementation, the stage transition index can be determined according to the following formula:

[0046]

[0047]

[0048] in, Indicates the first The stage transition index corresponding to each sliding window is a dimensionless quantity. It is a smoothing mapping function used to compress the result of fusing multiple physical quantities into the range of 0 to 1, so as to avoid the trigger signal jump caused by the instantaneous noise of a single window; Indicates the first Average rate of humidity change within each window This represents the average rate of humidity change within the constant-rate dehydration stability range of the current batch. and These represent the variances of the humidity change rate between the current window and the constant-rate steady-state interval, respectively. and These represent the variance of pressure fluctuations in the current window and the constant-speed steady-state interval, respectively. The trend persistence coefficient, its physical meaning is the most recent trend. The degree to which the rate of humidity change in a window continues to decrease is indicated by the value being closer to 1, which shows that the continuous decay is more stable. The number of consecutive windows can be determined based on air turnover time, process cycle time, and historical batch statistics.

[0049] The above formula combines four types of information: "whether the humidity change rate decreases," "whether the humidity change rate fluctuation converges," "whether the pressure stabilizes," and "whether the downward trend continues," rather than relying on a single moment or a single parameter to make a switching judgment. Therefore, it is more able to capture the true dynamic transition from constant-rate dehydration to falling-rate dehydration. The controller does not... Instead of directly determining it as a fixed value, it is compared with the distribution of the current batch's stable sample band and historical turning point sample bands; when When a sample continuously enters the transition sample band and simultaneously leaves the stable sample band, the controller outputs a stage transition trigger signal. If the signal is confirmed within the window... If the sample size returns to a stable band, the triggering is canceled and monitoring continues. The controller can also... The data is written back to the state matrix as an auxiliary field for the next round of sliding window analysis. For batches with skewed distributions, the sample band can be replaced by quantile intervals, but this embodiment preferably uses a statistical band formed by the mean and standard deviation, which is convenient for on-site calibration and subsequent maintenance. To enhance real-time performance, the controller can set a confirmation window after the trigger window. If the humidity change rate continues to decrease and the pressure fluctuation continues to converge within the confirmation window, the stage transition trigger signal will take effect; if the state within the confirmation window returns to the stable sample band, it is considered a short-term disturbance and control switching will not be triggered. This can effectively suppress misjudgments caused by momentary disturbances such as door opening, material replenishment, and air duct movement.

[0050] S3. After responding to the stage transition trigger signal, the vegetable dehydration processing controller constructs a residual energy demand factor based on the degree of decay of the current humidity change rate and the average humidity change rate of the constant-speed dehydration stage, the degree of pressure fluctuation stabilization, the stage transition index, and the soft coupling relationship between temperature deviation. This factor simultaneously determines the coordinated reduction target of heating power and fan speed, forming a dynamic energy input strategy.

[0051] Specifically, the controller reads the mean humidity change rate of the current window, the mean humidity change rate of the current batch constant-rate dehydration, the variance of the current window pressure fluctuation, the variance of the current batch constant-rate dehydration pressure fluctuation, and the current chamber temperature from the state matrix and baseline cache. Preferably, the temperature is formed by the arithmetic mean of three evenly distributed temperature measuring points in the upper, middle, and lower parts of the cavity; when one of the temperature measuring points fails, the controller uses the average of the remaining two measuring points and combines it with the historical offset of that measuring point for compensation. The target temperature for the current stage is constant and does not change with a single control cycle. For example, for leafy vegetables, a target temperature of around 65°C can be preferred in this stage, and the temperature should be maintained within the allowable deviation range for stable operation. The ratio of the humidity change rate to the constant-speed baseline reflects the remaining proportion of the current moisture migration capacity relative to the constant-speed stage; the pressure fluctuation variance reflects whether the airflow load and steam release are still at a high level; the cavity temperature reflects whether the actual heating has approached the target process temperature; and the stage transition index characterizes the confidence level of the transition from constant speed to deceleration. The controller does not simply reduce power and fan speed simultaneously in a linear proportion, but rather links the four types of signals through a residual energy demand factor, so that the decrease in heating power and fan speed can both follow the decay of the dehydration rate and avoid damaging the dehumidification capacity due to excessive reduction. The practical significance of the control strategy is that when a large amount of free water on the surface of the vegetables has evaporated and the migration of bound water inside has become the dominant process, continuing to maintain a high heat flow and high air volume in the constant-rate stage will only increase energy consumption without significantly improving the dehydration rate. Therefore, the energy input must be reduced with the change of state; however, the reduction cannot be too large at once, otherwise it will lead to a sudden drop in cavity temperature, accumulation of moisture in the air duct, or local dampness in the material layer. Therefore, it is necessary to smooth the decay through the residual energy demand factor.

[0052] In one implementation, the dynamic energy input strategy can be determined according to the following formula:

[0053]

[0054]

[0055]

[0056] in, The residual energy demand factor is a dimensionless quantity, and its physical meaning is the ratio of heating and air supply intensity that still needs to be maintained under the current dehydration state. and These represent the average rate of humidity change in the current window and the constant-speed steady-state interval, respectively. and These represent the variance of pressure fluctuations in the current window and the constant-rate steady-state interval, respectively. The stage transition index obtained from S2; This is a smoothing mapping function used to compress the cavity temperature deviation to the range of 0 to 1, avoiding power jumps caused by slight temperature fluctuations; The target chamber temperature for the current batch at this dehydration stage is derived from the process formulation or the temperature baseline of historical qualified batches. This refers to the current cavity temperature. The allowable temperature deviation range is determined by the thermal inertia of the heating unit, product quality requirements, and cavity temperature uniformity. It is a dimensionless coefficient and can be calibrated according to vegetable type, material layer thickness, equipment model and historical energy saving effect; and These represent the average heating power and average fan speed within the constant-rate dehydration stability range of the current batch, respectively. and These represent the minimum stable operating boundaries of the heating unit and the fan under the current process, respectively, derived from the equipment performance curve, minimum effective dehumidification capacity, and safe operating boundary.

[0057] The core of this formula is multiplicative coupling rather than simple linear scaling: when the rate of change in humidity decreases, pressure fluctuations stabilize, the phase transition index increases, and the temperature deviation converges towards the target... The power and speed will decrease synchronously, converging towards the low-load range. When the cavity temperature is low or the pressure is not yet stable, the temperature and pressure terms will suppress excessively rapid power reduction, allowing the system to retain the necessary hot air delivery capacity. The heating power and fan speed do not decrease proportionally; instead, different exponential parameters create different response sensitivities. Typically, the heating power is more sensitive to humidity migration attenuation, while the fan speed maintains a smoother decline curve to avoid humid air retention within the material layer. To make this control process directly implementable, the controller can recalculate within each control cycle. and based on The changing trend determines the power and speed targets for the next cycle; if A continuous decrease indicates a stable transition from the dehydration state to the deceleration phase, and the controller continues to adjust the power and speed according to the predetermined decreasing curve; if If the temperature rebounds in a short period of time, it indicates that the humidity or pressure of the air outlet has not yet stabilized. The controller can pause further power reduction and prioritize maintaining the fan speed or moderately increasing the heating power to restore the dehumidification capacity.

[0058] In one calibration embodiment, using cabbage as the object, an L16 orthogonal experiment was conducted. The evaluation index was a weighted sum of unit energy consumption, final product moisture content deviation, cavity temperature fluctuation, and stage switching lag time, which yielded a set of feasible parameters, for example... .in, Used to enhance the suppression effect of pressure disturbances on the derivative of humidity. Used to balance humidity decay, humidity fluctuation, pressure convergence, and trend persistence. The sensitivity used to control the decline in residual energy demand. This is used to make the decrease curve of the fan speed slightly flatter than that of the heating power. Used to limit the compressibility of the stage transition index with respect to the energy factor. This is used to correct the energy input for temperature deviation. This embodiment is only for illustrating the parameter calibration method and does not limit the scope of protection; for other vegetable varieties, those skilled in the art can obtain the coefficient combination for the corresponding variety simply by using the same experimental design and evaluation indicators, without any creative effort. If the tissue density, moisture content structure, or material layer resistance of a certain vegetable variety is significantly different from that of cabbage, the orthogonal experimental level can be adjusted within the same framework without redesigning the entire control logic. The core here is not a specific value, but rather the multiplicative coupling of four factors—humidity decay, pressure convergence, trend persistence, and temperature deviation—to bind dynamic operating conditions with coordinated control of heating and air supply, thereby avoiding the ineffective energy consumption caused by traditional fixed power or fixed wind speed strategies during the dehydration phase.

[0059] S4. Finally, the dynamic energy input strategy is converted into control quantities for the heating unit and the fan actuator and executed. Simultaneously, during execution, actual power, actual speed, outlet humidity, and pressure fluctuations are continuously collected and written back to the state matrix, ensuring that the heating power and fan speed always maintain operating parameters that match the current dehydration rate, thus completing the energy-saving vegetable dehydration processing control. Specifically, the controller first determines the appropriate parameters based on the heating unit type... Converted into executable duty cycle, thyristor conduction ratio, steam valve opening, or fuel supply; then, based on the fan drive method... This is converted into inverter frequency setpoint, speed closed-loop target, or motor torque command. In actual engineering, these conversion relationships are preferably derived from the equipment's factory calibration curves, PLC control protocols, and on-site commissioning results, rather than directly forcing power and speed values ​​onto the actuator. When issuing control quantities, the controller preferably uses a ramp transition rather than a single-cycle jump to avoid overshoot caused by the thermal inertia of the heating unit and the mechanical inertia of the fan. When the actuator can only accept discrete speeds, the controller can select the closest speed and maintain the overall decreasing trend without reversing it. In this way, even if the target value is updated in each control cycle, the actual execution path remains continuous and smooth, without fluctuations in cavity temperature or duct pressure caused by frequent jitter. For different actuators, the controller can establish a device parameter mapping table. For example, the electric heating unit corresponds to the "duty cycle-output power" curve, the variable frequency fan corresponds to the "frequency-speed" curve, and the steam heat exchanger corresponds to the "valve opening-heat input" curve. The initial curve is obtained through no-load calibration and standard operating condition calibration before the equipment is put into operation, and then the curve is locally corrected based on the feedback data during operation.

[0060] Furthermore, during execution, the controller continuously reads the actual output power of the heating unit, the actual speed of the fan, the humidity at the air outlet, and the pressure in the processing chamber, and writes these feedback results back to the state matrix as input data for the next rounds S1 and S2. If there is a deviation between the actual output and the target output, the controller prioritizes compensation based on the response lag of the actuator, rather than immediately changing the overall control logic; if the deviation comes from changes in raw material loading, changes in duct resistance, or sensor reading drift, the controller corrects the control direction in subsequent windows by updating the baseline and adjusting the residual energy demand factor. In other words, S4 is not a one-time output control command, but rather truly implements the state identification and energy regulation closed loop formed from S1 to S3 at the execution layer, and then feeds back the actual operating results after execution to the previous round's state matrix, so that the next round's stage identification and heating control continue to converge based on the latest operating conditions. The batch operation record can also store the constant speed stable range, stage transition critical point, dynamic energy input strategy, actual energy consumption, and final product test results for the current batch. Only when the final product moisture content is qualified and the unit energy consumption meets the standard will the corresponding transition window statistics be written into the historical sample library to update the transition sample band. Unqualified batch data is marked as abnormal and removed, and does not participate in the sample band update, thus ensuring that hot air heating and fan air delivery maintain a dynamic matching state throughout the dehydration process of different batches of vegetables. Furthermore, to ensure the traceability of the historical sample library, the controller can store the category, loading range, initial moisture content range, stage transition timestamp, etc., corresponding to each batch of qualified samples. Trajectory, The trajectory and final product test results are written to the database together. This allows subsequent batches to be retrieved by category and operating condition when updating sample bands, avoiding the mixing of data from different vegetable types. For restart conditions after equipment maintenance, duct cleaning, or sensor replacement, the system can perform a short-cycle initialization calibration to bring stable and transitional sample bands back to the new equipment state and re-converge, ensuring that the control strategy remains effective during long-term operation.

[0061] In summary, the energy-saving vegetable dehydration processing control method based on the embodiments of the present invention is explained. It simultaneously acquires information on humidity changes at the air outlet and pressure fluctuations in the processing chamber during the vegetable dehydration process, constructing a dehydration dynamic state matrix that simultaneously reflects the strength of moisture migration and the state of airflow disturbance. Then, based on this state matrix, it identifies the critical point of transition from a constant-rate dehydration stage to a deceleration stage and generates a stage transition trigger signal. Upon responding to the trigger signal, it coordinates and gradually reduces the heating power and fan speed according to the degree of attenuation of the current dehydration rate relative to the constant-rate stage, forming a dynamic energy input strategy and driving the actuator to ensure that the heating power and fan speed are matched with the current dehydration rate in real time. This overcomes the technical defects of existing technologies where the fixed-sequence switching strategy cannot accurately capture the real-time transition points of dehydration dynamic characteristics, resulting in a severe mismatch between energy input and the actual dehydration rate, effectively improving the energy utilization efficiency of the vegetable dehydration processing process.

[0062] Here, those skilled in the art will understand that the specific operations of each step in the above-mentioned energy-saving vegetable dehydration processing control system have been referenced above. Figures 1 to 3 The energy-saving vegetable dehydration processing control method is described in detail in the description of the method, and therefore, its repeated description will be omitted.

[0063] According to another aspect of the present invention, an energy-saving vegetable dehydration processing control system is provided, comprising:

[0064] The state perception and feature extraction module is used to acquire humidity change information and pressure fluctuation information during the vegetable dehydration process, and to construct a dehydration dynamic state matrix based on the humidity change information and the pressure fluctuation information.

[0065] The stage transition identification module is used to identify the critical point of dehydration stage transition based on the dehydration kinetic state matrix and generate a stage transition trigger signal.

[0066] The dynamic energy control module is used to respond to the stage transition trigger signal and, based on the degree of decay of the current dehydration rate relative to the dehydration rate of the constant-rate dehydration stage, to coordinately reduce the heating power and the fan speed to obtain a dynamic energy input strategy.

[0067] The execution and self-optimization module is used to drive the heating unit and the fan actuator according to the dynamic energy input strategy so that the heating power and fan speed are matched with the current dehydration rate.

[0068] In summary, the energy-saving vegetable dehydration processing control system based on the embodiments of the present invention is explained. It simultaneously acquires information on humidity changes at the air outlet and pressure fluctuations in the processing chamber during the vegetable dehydration process, constructing a dehydration dynamic state matrix that simultaneously reflects the strength of moisture migration and the state of airflow disturbance. Then, based on this state matrix, it identifies the critical point of transition from a constant-speed dehydration stage to a deceleration stage and generates a stage transition trigger signal. Upon responding to the trigger signal, it coordinates and gradually reduces the heating power and fan speed according to the degree of attenuation of the current dehydration rate relative to the constant-speed stage, forming a dynamic energy input strategy and driving the actuator to ensure that the heating power and fan speed are matched with the current dehydration rate in real time. This overcomes the technical defects of existing technologies where the fixed-sequence switching strategy cannot accurately capture the real-time transition points of dehydration dynamic characteristics, resulting in a severe mismatch between energy input and the actual dehydration rate, effectively improving the energy utilization efficiency of the vegetable dehydration processing process.

Claims

1. An energy-saving vegetable dehydration processing control method, characterized in that, include: Obtain humidity change information and pressure fluctuation information during the vegetable dehydration process, and construct a dehydration dynamic state matrix based on the humidity change information and the pressure fluctuation information; Based on the aforementioned dehydration kinetic state matrix, the critical point for the transition between dehydration stages is identified, and a stage transition trigger signal is generated; In response to the stage transition trigger signal, the heating power and fan speed are coordinated and gradually reduced according to the degree of decay of the current dehydration rate relative to the dehydration rate of the constant rate dehydration stage, so as to obtain a dynamic energy input strategy. The heating unit and fan actuator are driven according to the dynamic energy input strategy to match the heating power and fan speed with the current dehydration rate.

2. The energy-saving vegetable dehydration processing control method according to claim 1, characterized in that, The humidity change information includes an air outlet humidity change sequence and a humidity change rate sequence obtained from the air outlet humidity change sequence; the pressure fluctuation information includes a processing cavity pressure fluctuation sequence. The humidity change rate sequence is obtained by pressure-weighted local regression differentiation. The pressure-weighted local regression differentiation includes: determining the degree of deviation between the pressure fluctuation corresponding to each humidity sampling point and the average pressure fluctuation in the local time window; determining the regression weight of each humidity sampling point according to the degree of deviation; and calculating the local regression slope of the air outlet humidity change sequence based on the regression weight.

3. The energy-saving vegetable dehydration processing control method according to claim 2, characterized in that, The dehydration kinetic state matrix includes sampling time, outlet humidity content, humidity change rate, processing chamber pressure fluctuation, humidity change rate variance, pressure fluctuation variance, and data integrity markers. The continuous time window used to construct the dehydration kinetic state matrix is ​​jointly determined by the air turnover cycle of the hot air circulation system, sensor response time, and control cycle, and is rolled over and corrected based on the stage transition backtracking results after batch processing.

4. The energy-saving vegetable dehydration processing control method according to claim 3, characterized in that, The critical point for the dehydration stage transition is determined by the stage transition index, which includes a humidity change rate decay term, a humidity change rate variance convergence term, a pressure fluctuation variance convergence term, and a trend persistence term. The humidity change rate decay term is used to characterize the degree of decrease of the mean humidity change rate of the current window relative to the mean humidity change rate of the constant rate dehydration stable interval. The humidity change rate variance convergence term is used to characterize the degree of convergence of the variance of the humidity change rate of the current window relative to the variance of the humidity change rate of the constant rate dehydration stable interval. The pressure fluctuation variance convergence term is used to characterize the degree of convergence of the variance of the pressure fluctuation of the current window relative to the variance of the pressure fluctuation of the constant rate dehydration stable interval. The trend persistence term is used to characterize the persistence of the humidity change rate evolving in a downward direction within multiple consecutive windows.

5. The energy-saving vegetable dehydration processing control method according to claim 4, characterized in that, The stage transition index corresponds to a stable sample band and a transition sample band; The stable sample band includes the stage transition index statistical interval of the constant rate dehydration stable window confirmed by offline moisture content in historical qualified batches, and the turning sample band includes the stage transition index statistical interval of the stage transition window confirmed by offline moisture content in historical qualified batches. When the stage transition index corresponding to the current window continuously enters the turning sample band and leaves the stable sample band, the stage transition trigger signal is generated; wherein, only batches with qualified final product moisture content and unit energy consumption are used to update the stable sample band and the turning sample band.

6. The energy-saving vegetable dehydration processing control method according to claim 5, characterized in that, The dynamic energy input strategy includes residual energy demand factor, target heating power and target fan speed; The residual energy demand factor is obtained by coupling the ratio of the current humidity change rate to the mean humidity change rate during the constant-rate dehydration stage, the ratio of the current pressure fluctuation variance to the pressure fluctuation variance during the constant-rate dehydration stage, the stage transition index, and the cavity temperature deviation. The target heating power is determined based on the residual energy demand factor and the minimum stable operating boundary of the heating unit, and the target fan speed is determined based on the residual energy demand factor and the minimum effective dehumidification operating boundary of the fan. The target heating power and the target fan speed have different decreasing response sensitivities.

7. The energy-saving vegetable dehydration processing control method according to claim 6, characterized in that, The cavity temperature deviation is determined by the current cavity temperature and the target cavity temperature; the current cavity temperature includes the average temperature of the upper, middle and lower temperature measuring points inside the processing cavity; When any temperature measuring point fails, the current cavity temperature is obtained by compensating for the average temperature of the remaining temperature measuring points with the historical offset of the failed temperature measuring point; the target cavity temperature is the constant process temperature corresponding to the current dehydration stage.

8. An energy-saving vegetable dehydration processing control system, characterized in that, include: The state perception and feature extraction module is used to acquire humidity change information and pressure fluctuation information during the vegetable dehydration process, and to construct a dehydration dynamic state matrix based on the humidity change information and the pressure fluctuation information. The stage transition identification module is used to identify the critical point of dehydration stage transition based on the dehydration kinetic state matrix and generate a stage transition trigger signal. The dynamic energy control module is used to respond to the stage transition trigger signal and, based on the degree of decay of the current dehydration rate relative to the dehydration rate of the constant-rate dehydration stage, to coordinately reduce the heating power and the fan speed to obtain a dynamic energy input strategy. The execution and self-optimization module is used to drive the heating unit and the fan actuator according to the dynamic energy input strategy so that the heating power and fan speed are matched with the current dehydration rate.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the energy-saving vegetable dehydration processing control method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the energy-saving vegetable dehydration processing control method according to any one of claims 1 to 7.

Images