An ice-making vending machine adaptive energy-saving control method and system

Abstract

The application discloses a kind of ice-making vending machine adaptive energy-saving control method and system, the method includes: obtaining the historical sales data of ice-making vending machine, the environmental temperature prediction data of future period and date attribute information, and input trained demand prediction model, obtain the ice demand prediction curve in future control time domain;Ice-making energy efficiency model of ice-making vending machine is obtained;Current ice storage is obtained;With the minimum total energy consumption in future control time domain as target, the ice demand prediction curve, current ice storage and ice-making energy efficiency model are used as input, and the ice-making power sequence is solved under the ice storage constraint condition Rolling optimization is obtained Optimal ice-making power sequence;The ice-making power instruction corresponding to current control period is extracted and the ice maker is controlled to run.The application can reduce the overall operation energy consumption of ice-making vending machine under the premise of meeting the ice demand, improve the adaptability of equipment to demand change and environmental change.

Description

Technical Field

[0001] This invention belongs to the field of refrigeration equipment control technology, specifically relating to an adaptive energy-saving control method and system for ice vending machines. Background Technology

[0002] An ice vending machine is a terminal device that integrates ice making, ice storage, and automatic vending functions, and is widely used in commercial districts, scenic spots, convenience stores, and other places. Its main energy consumption comes from the operation of refrigeration components such as the ice compressor and condenser fan. Because ice demand varies significantly across time periods, if the ice maker always operates at a fixed power level, it will lead to excessive ice storage during off-peak periods, resulting in the natural dissipation and waste of cold energy, while during peak periods, insufficient capacity may affect sales.

[0003] To address these issues, existing methods typically involve collecting historical sales data, real-time ambient temperature, and holiday information to predict future ice demand and adjust ice-making strategies accordingly. For example, low-frequency intermittent ice making or stopping ice making during off-peak periods, and high-frequency ice making to build up cooling capacity reserves before peak demand arrives. While this strategy avoids the waste of operating at full capacity throughout the day and achieves peak shaving to some extent, it usually involves directly starting and stopping the compressor based on demand forecasts or switching between a limited number of fixed power settings.

[0004] This control method ignores a crucial physical characteristic of ice makers: their unit ice production energy consumption is not constant but exhibits a significant non-linear characteristic that varies with actual operating power and ambient temperature. Specifically, the amount of cooling produced per kilowatt-hour of electricity consumed varies significantly depending on the compressor's operating frequency. Simply operating ice makers at low frequencies during off-peak hours and high frequencies during peak hours only considers demand matching over time, failing to utilize the energy efficiency differences of the ice maker under different operating conditions. This may result in the ice maker operating at high power and low energy efficiency for extended periods during peak hours, achieving the ice production task but incurring unnecessary additional energy costs.

[0005] Therefore, in order to address the above-mentioned technical problems, it is necessary to provide an adaptive energy-saving control method and system for ice vending machines. Summary of the Invention

[0006] The purpose of this invention is to provide an adaptive energy-saving control method and system for ice vending machines, which can solve the technical problems mentioned in the background art.

[0007] To achieve the above objectives, a specific embodiment of the present invention provides the following technical solution:

[0008] An adaptive energy-saving control method for an ice vending machine includes the following steps: S1, acquiring historical sales data, predicted ambient temperature data for future time periods, and date attribute information of the ice vending machine; inputting the historical sales data, predicted ambient temperature data, and date attribute information into a trained demand prediction model to obtain a predicted ice demand curve in the future control time domain; S2, acquiring the ice-making energy efficiency model of the ice vending machine, wherein the ice-making energy efficiency model is a model of the mapping relationship between ice-making power and ice-making speed under different ambient temperatures, and under a fixed ambient temperature, the unit ice production energy consumption of the ice vending machine exhibits a nonlinear characteristic as the ice-making power changes. S3. Real-time collection of ice storage data in the ice vending machine's ice storage compartment to obtain the current ice storage volume; S4. With the goal of minimizing total energy consumption in the future control time domain, the ice demand prediction curve, current ice storage volume, and ice-making energy efficiency model are used as inputs. Under the preset ice storage volume constraint, the ice-making power sequence is subjected to rolling optimization to obtain an optimal ice-making power sequence that meets the ice demand requirement and has the lowest total energy consumption; S5. Ice-making power instructions corresponding to the current control cycle are extracted from the optimal ice-making power sequence, and the ice maker is controlled to operate according to the ice-making power instructions.

[0009] In one or more embodiments of the present invention, the demand prediction model in step S1 is a long short-term memory network model, and the ice demand prediction curve consists of a set of ice demand data for each discrete time period in the future, arranged in chronological order.

[0010] In one or more embodiments of the present invention, the ice-making energy efficiency model described in step S2 is obtained by calibrating the ice-making speed of the ice maker under different ambient temperatures and different operating power in advance, and establishing a two-dimensional mapping relationship with ambient temperature and ice-making power as inputs and ice-making speed as output.

[0011] In one or more embodiments of the present invention, the rolling optimization solution of the ice-making power sequence in step S4 specifically includes: generating multiple candidate ice-making power sequences with the current ice storage as the initial value of the state variable; for each candidate ice-making power sequence, determining the predicted ice-making speed for each control cycle based on the ice-making energy efficiency model and the ambient temperature of the corresponding time period, and iteratively calculating the predicted ice storage at the end of each control cycle in the future control time domain in conjunction with the ice demand prediction curve, forming a trajectory of ice storage change; selecting the candidate ice-making power sequence that minimizes the total predicted energy consumption from all candidate ice-making power sequences that satisfy the ice storage constraint as the optimal ice-making power sequence, wherein the ice storage constraint includes the ice storage not exceeding the upper limit of the ice storage capacity and not lower than the lower limit of the safety stock.

[0012] In one or more embodiments of the present invention, when calculating the total predicted energy consumption, a time-of-use electricity price factor is introduced, so that the optimal ice-making power sequence is the ice-making power sequence with the lowest total predicted electricity cost.

[0013] In one or more embodiments of the present invention, step S5, which controls the operation of the ice maker according to the ice-making power command, specifically involves converting the ice-making power command into a variable frequency drive signal for the compressor to continuously adjust the operating frequency of the compressor.

[0014] In one or more embodiments of the present invention, the method executes steps S1 to S5 in a fixed control cycle. Each time it is executed, step S4 outputs only the optimal ice-making power sequence in the future control time domain, while step S5 executes only the first ice-making power command in the optimal ice-making power sequence. In the next control cycle, the entire optimization process is repeated based on the re-acquired current ice storage and the newly acquired ice demand prediction curve.

[0015] In one or more embodiments of the present invention, the mapping relationship represented by the ice-making energy efficiency model enables the optimal ice-making power sequence to proactively avoid high-power, low-efficiency operating ranges and prioritize the use of the most energy-efficient power ranges for advance ice storage, so as to achieve the lowest total energy consumption while meeting ice demand.

[0016] To achieve the above objectives, the present invention also provides an adaptive energy-saving control system for an ice vending machine, comprising a demand prediction module, an energy efficiency model storage module, an ice storage detection module, an optimization control module, and an execution drive module. The demand prediction module acquires historical sales data, future ambient temperature prediction data, and date attribute information from the ice vending machine. It inputs the historical sales data, ambient temperature prediction data, and date attribute information into a trained demand prediction model to obtain a predicted ice demand curve for the future control time domain. The energy efficiency model storage module stores the ice-making energy efficiency model of the ice vending machine. This model represents the mapping relationship between ice-making power and ice-making speed under different ambient temperatures. Furthermore, under any fixed ambient temperature, the ice-making energy efficiency ratio first increases and then decreases with increasing ice-making power. The nonlinear characteristics of the ice storage capacity are analyzed. The ice storage capacity detection module is used to collect the ice storage capacity data in the ice storage of the ice vending machine in real time to obtain the current ice storage capacity. The optimization control module is connected to the demand prediction module, the energy efficiency model storage module and the ice storage capacity detection module respectively. It is used to minimize the total energy consumption in the future control time domain, and take the ice demand prediction curve, the current ice storage capacity and the ice making energy efficiency model as input. Under the preset ice storage capacity constraint, it performs rolling optimization to solve the ice making power sequence to obtain an optimal ice making power sequence that meets the ice demand requirement and has the lowest total energy consumption. The execution drive module is connected to the optimization control module and the ice maker of the ice vending machine. It is used to extract the ice making power command corresponding to the current control cycle from the optimal ice making power sequence and control the ice maker to operate according to the ice making power command.

[0017] In one or more embodiments of the present invention, the optimization control module is specifically used to: generate multiple candidate ice-making power sequences with the current ice storage as the initial value of the state variable; for each candidate ice-making power sequence, determine the predicted ice-making speed for each control cycle according to the ice-making energy efficiency model and the ambient temperature of the corresponding time period, and iteratively calculate the predicted ice storage at the end of each control cycle in the future control time domain by combining the ice demand prediction curve, forming an ice storage change trajectory; among all candidate ice-making power sequences that meet the ice storage constraint, select the candidate ice-making power sequence with the minimum total predicted energy consumption as the optimal ice-making power sequence, wherein the ice storage constraint includes the ice storage not exceeding the upper limit of the ice storage capacity and not lower than the safety stock limit.

[0018] Compared with the prior art, the beneficial effects of the present invention are as follows: under the premise of meeting the ice demand and ice storage safety constraints, it can actively avoid the high power and low energy efficiency operating range, and give priority to using the target power range with the best or relatively better energy efficiency for early ice storage, thereby obtaining an operating path with lower total energy consumption or lower total electricity cost, reducing unnecessary power consumption, and improving the operating economy and energy-saving effect of the ice vending machine. Attached Figure Description

[0019] 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 only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a flowchart of an adaptive energy-saving control method for an ice vending machine according to an embodiment of the present invention;

[0021] Figure 2 This is a flowchart of ice demand forecasting in one embodiment of the present invention;

[0022] Figure 3 This is a flowchart illustrating the process of obtaining the ice-making energy efficiency model in one embodiment of the present invention;

[0023] Figure 4 This is a flowchart of the rolling optimization solution in one embodiment of the present invention;

[0024] Figure 5 This is a flowchart illustrating the execution drive control process in one embodiment of the present invention.

[0025] Figure 6 This is a flowchart of the rolling closed-loop control in one embodiment of the present invention;

[0026] Figure 7 This is a structural block diagram of an adaptive energy-saving control system for an ice vending machine according to an embodiment of the present invention.

[0027] Explanation of key figure labels:

[0028] 101. Demand forecasting module; 102. Energy efficiency model storage module; 103. Ice storage detection module; 104. Optimization control module; 105. Execution drive module. Detailed Implementation

[0029] To enable those skilled in the art to better understand the technical solutions in this disclosure, the technical solutions in the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this disclosure, and not all embodiments. Based on the embodiments in this disclosure, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this disclosure.

[0030] An adaptive energy-saving control method for an ice vending machine according to one embodiment of the present invention can avoid high-power, low-energy-efficiency operating ranges while meeting ice demand, and prioritize the use of power ranges with better energy efficiency for pre-storing ice, thereby reducing the overall operating energy consumption of the ice vending machine.

[0031] like Figure 1 As shown, the adaptive energy-saving control method for ice vending machines includes steps S1 to S5, which are executed in a rolling manner with a fixed control cycle. The control cycle can be set according to the ice-making capacity, ice storage capacity, and vending frequency of the ice vending machine, for example, 5 minutes, 10 minutes, 15 minutes, or 30 minutes. Generally, the control cycle is set to 15 minutes, and the future control time domain is set to 6 hours, then the future control time domain includes 24 discrete control cycles.

[0032] Within a control cycle, historical data collection, demand forecasting, status updates, rolling optimization, and instruction issuance are completed sequentially, and the above process is repeated in the next control cycle.

[0033] like Figure 2 As shown, S1: Obtain historical sales data, future ambient temperature prediction data, and date attribute information of the ice vending machine; input the historical sales data, ambient temperature prediction data, and date attribute information into the trained demand prediction model to obtain the ice demand prediction curve in the future control time domain.

[0034] When a new control cycle begins, the system first acquires historical sales data from the ice vending machine, recorded cyclically over several consecutive days, weeks, or months. This historical sales data can include information such as the quality of ice sold within each control cycle, the number of transactions, the sales time, replenishment records, and ice shortage records. The quality of the ice sold can serve as a direct indicator of ice demand, or it can be adjusted by combining data such as ice shortage records and canceled orders to arrive at the actual ice demand.

[0035] Simultaneously acquire ambient temperature forecast data for the future control time domain. Ambient temperature forecast data can originate from weather forecast interfaces, cloud-based meteorological services, forecasts obtained from local temperature sensors combined with trend prediction algorithms, or a combination of these methods. The ambient temperature forecast data is then organized into average or representative temperature values ​​for each discrete time period according to the control cycle.

[0036] Date attribute information is used to characterize the date context of each time period within the future control time domain. It can include weekday information, weekday or holiday identifiers, seasonal information, month information, special event identifiers, and business district customer flow level identifiers. For scenarios where ice demand is greatly affected by pedestrian flow, such as scenic spots, commercial streets, and farmers' markets, date attribute information can improve the ability of demand forecasting models to identify regular peak and trough changes.

[0037] The aforementioned historical sales data, ambient temperature forecast data, and date attribute information, after data cleaning, missing value handling, outlier correction, and normalization, are combined into a feature vector sequence and input into a trained demand forecasting model. The demand forecasting model can be a Long Short-Term Memory (LSTM) network model, a Recurrent Neural Network (RNN) model, a Temporal Convolutional Network (TCN) model, a Gradient Boosting Tree (GPRS) model, a Random Forest (RFR) model, or other time series forecasting models.

[0038] Preferably, the demand forecasting model employs a Long Short-Term Memory (LSTM) network model. This LSM network model includes an input layer, at least one LSTM hidden layer, and a fully connected output layer. During training, historical sales data, corresponding time-period ambient temperature data, and date attribute information are used as input samples, and the actual ice demand is used as a label for supervised learning. After the model training is complete, during the runtime phase, the model receives the currently constructed feature vector sequence and outputs the predicted ice demand for each discrete time period within the future control time domain.

[0039] The ice demand forecast curve can be represented as follows:

[0040]

[0041] in, This represents the predicted ice demand for the k-th control period within the future control time domain, where N represents the number of control periods included in the future control time domain. The unit for predicted ice demand can be kilograms.

[0042] To adapt the demand forecasting model to changes in ice usage habits, seasonal variations, and sales scenarios, the parameters of the demand forecasting model can be periodically updated using newly accumulated actual sales data.

[0043] like Figure 3 As shown, S2, obtain the ice-making energy efficiency model of the ice vending machine.

[0044] The ice-making energy efficiency model is a mapping relationship between ice-making power and ice-making speed under different ambient temperatures. It describes the amount of ice a machine can produce per unit time under given ambient temperature and ice-making power conditions. This model allows for further calculation of unit ice production energy consumption or ice-making energy efficiency indicators, thereby identifying high-efficiency operating ranges, low-efficiency operating ranges, and target power ranges under different ambient temperatures.

[0045] The ice-making energy efficiency model can be expressed as:

[0046]

[0047] in, Indicates the ice-making speed. Indicates ice-making power. Indicates ambient temperature. This represents the mapping function obtained through experimental calibration or training with running data.

[0048] Ice-making energy efficiency models can be obtained through pre-calibration. The calibration process can be completed before the equipment leaves the factory, during installation and commissioning, or during maintenance and calibration. Specifically, the ice vending machine is placed under different ambient temperature conditions, such as 15℃, 20℃, 25℃, 30℃, 35℃, and 40℃. After each ambient temperature stabilizes, the compressor is controlled to operate at multiple different ice-making power points, and the mass of ice produced per unit time at each power point is recorded, thus obtaining multiple sets of experimental data on the relationship between ambient temperature, ice-making power, and ice-making speed. Subsequently, the experimental data is filtered, interpolated, and smoothed to establish a two-dimensional lookup table, a piecewise interpolation model, or a fitting model.

[0049] When the ice-making energy efficiency model is stored in the form of a two-dimensional lookup table, one dimension of the lookup table is the ambient temperature, the other dimension is the ice-making power, and the table elements are the corresponding ice-making speeds. When the actual ambient temperature or ice-making power falls between two calibration points, the corresponding ice-making speed can be obtained using linear interpolation, bilinear interpolation, or spline interpolation.

[0050] As an alternative approach, the ice-making energy efficiency model can also be represented using a regression function. To ensure the model reflects the nonlinear characteristics of ice-making energy efficiency as a function of ice-making power, the ice-making rate can be expressed as:

[0051]

[0052] in, , , The coefficient function is related to ambient temperature and can be obtained by fitting experimental calibration data. , , It can be represented by a linear function, a quadratic function, a piecewise function, or a lookup table.

[0053] For example, , , We can use different methods regarding A linear function, in which case the ice-making energy efficiency model can be specifically expressed as:

[0054]

[0055] in, , , , , , These are the model coefficients obtained by fitting the calibration data.

[0056] The above function form is only an example. In practical applications, higher-order polynomials, piecewise functions, exponential models, BP neural network models, or other computable models that can reflect the relationship between ice-making speed and changes in ambient temperature and ice-making power can also be used.

[0057] Based on ice-making speed and ice-making power This allows for further calculation of the ice-making energy efficiency ratio. Energy consumption per unit of ice production The energy efficiency ratio for ice making can be expressed as:

[0058]

[0059] The energy consumption per unit of ice production can be expressed as:

[0060]

[0061] in, The larger the value, the greater the ice production per unit of electrical energy. The smaller the value, the lower the electrical energy required to produce a unit of ice. Because... Follow and It exhibits nonlinear variation. and It also varies with the ice-making power and ambient temperature. Under a fixed ambient temperature, the unit ice production energy consumption of an ice maker exhibits a non-linear characteristic as the ice-making power changes, and there can be a target power range that optimizes or relatively optimizes energy efficiency.

[0062] This ice-making energy efficiency model can identify high-efficiency and low-efficiency operating ranges under different ambient temperatures, and incorporate this energy efficiency difference into the rolling optimization solution of the subsequent ice-making power sequence.

[0063] It should be noted that this scheme does not require the mapping relationship represented by the ice-making energy efficiency model to conform to a specific functional form or curve shape across the entire temperature and power range, nor does it limit it to a curve with only a single extreme point. For example, under different ambient temperatures, the unit ice production energy consumption may exhibit a single-valley curve, a monotonically decreasing curve, a curve with a flat interval, or other nonlinear forms as a function of ice-making power. The core focus of this scheme is that the unit ice production energy consumption of the ice maker varies with ambient temperature and ice-making power. This variation can be quantitatively applied through the ice-making energy efficiency model, thereby identifying and prioritizing the use of target power ranges with relatively low unit ice production energy consumption during the rolling optimization process, rather than blindly concentrating high-power ice production during peak periods.

[0064] S3. Collect the ice storage data in the ice storage compartment of the ice vending machine in real time to obtain the current ice storage volume.

[0065] The ice storage capacity can be detected using an ultrasonic level gauge. Specifically, the ultrasonic level gauge is installed at the top of the ice storage silo, emitting ultrasonic pulses vertically downwards and receiving the echo signals reflected back from the ice surface. The controller calculates the distance from the sensor to the ice surface based on the ultrasonic transit time, and then combines this with the height of the ice storage silo, the cross-sectional area of ​​the silo, and the average bulk density of the ice blocks to calculate the current ice storage capacity. .

[0066] Alternatively, a load cell can be used to detect the current ice storage volume. The load cell is installed at the bottom of the ice storage tank or on its supporting structure, and the current ice storage volume is obtained by detecting changes in the weight of the ice blocks inside the tank. To improve detection stability, the collected ice storage volume data can be filtered using moving average filtering, median filtering, or amplitude limiting filtering to reduce errors caused by sensor noise, uneven ice block accumulation, and impacts from the ice dispensing mechanism.

[0067] Current ice storage The initial values ​​of the state variables used in the rolling optimization solution describe the available ice reserves at the current moment. When the loss of ultrasonic echo, sudden change in weighing signal, abnormal sensor communication, or signal quality below a preset threshold is detected, the system can switch to an estimation mode based on the difference between ice production and sales in the most recent few cycles, and limit the operating power of the compressor to avoid the risk of ice overflow or ice shortage.

[0068] like Figure 4 As shown in S4, with the goal of minimizing the total energy consumption in the future control time domain, the ice demand prediction curve, the current ice storage amount, and the ice-making energy efficiency model are used as inputs. Under the preset ice storage amount constraint, the ice-making power sequence is solved by rolling optimization to obtain an optimal ice-making power sequence that meets the ice demand requirement and has the lowest total energy consumption.

[0069] Suppose that the future control time domain includes N control cycles, and the duration of each control cycle is... The ice-making power sequence to be solved can be expressed as:

[0070]

[0071] in, This represents the ice-making power during the k-th control cycle. Ice-making power Meet the device's permissible power range constraints:

[0072]

[0073] in, This is the maximum ice-making power allowed by the ice maker. When the value is 0, it indicates that the ice maker stops making ice during this control cycle; when... A value greater than 0 indicates that the ice maker is operating at the corresponding power level.

[0074] The optimization problem aims to minimize the total predicted energy consumption in the future control time domain. The objective function can be expressed as:

[0075]

[0076] in, This represents the total predicted energy consumption within the future control time domain.

[0077] The state transition relationship of ice storage capacity can be expressed as:

[0078]

[0079] in, This represents the predicted ice storage volume at the end of the k-th control cycle. This represents the predicted ice storage capacity at the start of the k-th control cycle. This indicates the ambient temperature corresponding to the kth control cycle. and ice-making power The predicted ice-making speed is as follows. This represents the predicted ice demand for the k-th control period.

[0080] Over the entire future control time domain, the predicted ice storage volume must satisfy the ice storage volume constraint condition:

[0081]

[0082] in, This indicates the minimum safety stock level, used to prevent sales disruptions due to insufficient ice storage. This indicates the upper limit of ice storage capacity, used to avoid excessive ice storage leading to insufficient space, decreased ice quality, or wasted cooling capacity.

[0083] The optimization control module performs rolling optimization to solve the ice-making power sequence based on the aforementioned objective function, state transition relationship, and constraints. The solution method can employ dynamic programming, particle swarm optimization, genetic algorithm, sequential quadratic programming, mixed integer programming, or other numerical optimization algorithms capable of solving constrained optimization problems.

[0084] Preferably, a candidate sequence search method based on particle swarm optimization is used for solving the problem. Specifically, each particle represents a candidate ice-making power sequence, and the particle's position vector corresponds to the ice-making power value in each control cycle within the future control time domain. The algorithm uses the total predicted energy consumption as the fitness function and the upper and lower limits of ice storage capacity as constraints. Through particle velocity and position updates, it searches for the candidate ice-making power sequence with the lowest objective function value in the feasible solution space.

[0085] To improve the solution speed, the optimal ice-making power sequence obtained in the previous control cycle can be shifted forward after removing the first executed power value, and a new power value can be added at the end as part of the initial population of the current control cycle, thereby achieving a hot start.

[0086] As seen in the optimization process described above, this scheme comprehensively compares the energy consumption results of different candidate ice-making power sequences over the entire future control time domain. For different candidate ice-making power sequences that meet the same predicted ice demand and ice storage constraints, if a candidate sequence can utilize the target power range to make ice in advance during periods of lower demand and reduce the high-power, low-efficiency operation time during peak demand periods, then this candidate sequence can usually obtain a lower objective function value. Driven by the objective function, the optimization algorithm will prioritize selecting the candidate ice-making power sequence with the lower total predicted energy consumption as the optimal ice-making power sequence. Thus, it is possible to achieve the control effect of actively avoiding high-power, low-efficiency operation ranges and prioritizing the use of the most efficient or relatively efficient power ranges for advance ice storage.

[0087] For example, in a commercial district during a summer afternoon, a peak ice demand is predicted to occur in two hours. Following a traditional demand-following strategy, ice makers might operate at near-maximum power after the peak demand arrives to compensate for the rapidly decreasing ice reserves. However, according to ice-making energy efficiency models, operating at near-maximum power at the current ambient temperature may correspond to high energy consumption per unit of ice production. Using the method of this invention, ice makers can operate at near-target power levels during periods of lower ice demand, gradually increasing ice reserves to a reasonable level. When the peak demand arrives, the high-power operation time can be reduced, utilizing the existing ice reserves to meet sales demand, thereby reducing the total predicted energy consumption over the future control time domain.

[0088] To accommodate time-of-use pricing in commercial settings, a time-of-use pricing factor can be introduced when calculating total predicted energy consumption. This ensures that the optimal ice-making power sequence is the ice-making power sequence that minimizes the total predicted electricity cost. The objective function can then be expressed as:

[0089]

[0090] in, This represents the electricity price corresponding to the k-th control period. Using this objective function, under the premise of satisfying ice storage constraints and ice demand, the ice-making power sequence with the lower total predicted electricity cost is prioritized. For example, ice production can be appropriately increased during off-peak hours and unnecessary high-power ice production can be reduced during peak hours, thereby lowering equipment operating costs. This time-of-use electricity price factor can be used alone or in conjunction with the ice-making energy efficiency model, simultaneously considering the impact of unit ice production energy consumption and electricity prices at different times on operating costs.

[0091] Furthermore, to reduce frequent compressor start-stops or drastic power fluctuations, a smoothing penalty term for power changes between adjacent cycles can be introduced into the objective function. The objective function can then be expressed as:

[0092]

[0093] in, This is a smoothing weighting coefficient. By setting this smoothing penalty term, the optimal ice-making power sequence can be made more stable, reducing sudden changes in compressor operating frequency, which helps improve equipment operating stability and extend compressor life. The time-of-use electricity price factor and the smoothing penalty term can be used alone or in combination to achieve a balance between energy saving, economy, and equipment protection.

[0094] like Figure 5 As shown, S5, extract the ice-making power command corresponding to the current control cycle from the optimal ice-making power sequence, and control the ice maker to operate according to the ice-making power command.

[0095] In step S4, the optimal ice-making power sequence in the future control time domain is obtained. Subsequently, instead of executing the entire sequence at once, only the ice-making power corresponding to the first control cycle is extracted. This serves as the ice-making power command for the current control cycle. This ice-making power command is then converted into a variable frequency drive signal for the compressor.

[0096] Specifically, the ice-making power command can be adjusted according to a preset power-frequency correspondence. Convert to compressor target operating frequency Subsequently, the inverter is controlled via 0-10V analog voltage signals, PWM signals, RS-485 communication signals, or Modbus RTU protocol commands to make the compressor operate at the target frequency. Auxiliary components such as condenser fans and water pumps can also be controlled in conjunction with the compressor's operating status, ambient temperature, condensing pressure, or evaporating temperature.

[0097] For example, when the current ice-making power command corresponds to a compressor operating frequency of 45Hz, a corresponding control signal is sent to the frequency converter to make the compressor operate stably near 45Hz. During actual operation, operating parameters such as compressor current, voltage, input power, evaporation temperature, condensation temperature, and ice output speed can be collected, and the actual ice-making power can be made close to the target ice-making power through closed-loop regulation.

[0098] This step converts the digital power command obtained from the rolling optimization solution into the actual operating state of the compressor, enabling the ice maker to move beyond traditional start-stop control or fixed-gear control and achieve continuous adjustment within the allowable power range. By continuously adjusting the compressor's operating frequency, the target power range can be matched more precisely, thereby reducing energy efficiency losses caused by coarse gear shifting.

[0099] like Figure 6 As shown, the method of the present invention executes steps S1 to S5 in a fixed control cycle. During each execution, step S4 outputs the optimal ice-making power sequence for the future control time domain, while step S5 executes only the first ice-making power command in this optimal sequence. When the next control cycle arrives, the current ice storage is re-acquired, the ice demand prediction curve and ambient temperature prediction data are re-acquired, and the entire optimization process is repeated.

[0100] This rolling optimization mechanism enhances the method's adaptability to uncertainties. For example, when actual ice sales suddenly exceed the predicted value, the current ice storage volume collected in the next control cycle will be lower than the original predicted ice storage volume trajectory. The optimization process will then use this latest ice storage volume as a starting point to resolve the ice-making power sequence, thereby increasing subsequent ice production in a timely manner. Conversely, when actual ice sales are lower than the predicted value, subsequent ice-making power can be reduced to avoid excessive ice storage and wasted cooling capacity. When the predicted ambient temperature changes, the ice-making energy efficiency model can be recalculated based on the updated ambient temperature, making the power sequence planning results more consistent with the latest operating conditions.

[0101] Furthermore, the ice-making energy efficiency model can be set as an updatable model. Specifically, during long-term operation, the correspondence between actual ice-making power, ambient temperature, operating time, and actual ice production can be recorded, and this operating data can be used to revise the original ice-making energy efficiency model. For example, when the compressor ages, the condenser accumulates dust, the heat exchange efficiency decreases, or the refrigerant state changes, the actual ice-making speed under the same ice-making power may deviate. Recursive least squares algorithms with forgetting factors, incremental learning algorithms, or periodic offline calibration methods can be used to update the two-dimensional mapping table or fitting coefficients in the ice-making energy efficiency model, enabling the model to reflect the true energy efficiency status of the equipment after long-term operation, thereby maintaining the long-term effectiveness of energy-saving control.

[0102] Furthermore, an anomaly protection mechanism can be set up. When anomalies occur in the ice storage detection, ambient temperature prediction data, demand forecasting model, or ice-making energy efficiency model, the system switches to a preset safety control strategy. For example, when demand forecasting data is temporarily unavailable, the ice-making power can be determined based on the average ice sales volume of the most recent control cycles, the safety stock limit, and the current ice storage volume; when the ice storage sensor malfunctions, the current ice storage volume can be estimated based on historical ice production and sales volumes, and the compressor operating power can be limited; when the ice-making energy efficiency model fails to be invoked, a preset conservative power curve can be used to avoid ice overflow, ice shortage, or abnormal compressor operation.

[0103] To achieve the above-mentioned adaptive energy-saving control method for ice vending machines, the present invention also provides an adaptive energy-saving control system for ice vending machines.

[0104] like Figure 7 As shown, the system includes a demand forecasting module 101, an energy efficiency model storage module 102, an ice storage detection module 103, an optimization control module 104, and an execution drive module 105.

[0105] The demand forecasting module 101 is used to execute step S1, obtain historical sales data of the ice vending machine, environmental temperature forecast data for future periods, and date attribute information, and input the historical sales data, environmental temperature forecast data, and date attribute information into the trained demand forecasting model, and output the ice demand forecast curve in the future control time domain.

[0106] The energy efficiency model storage module 102 is used to store and provide an ice-making energy efficiency model. The ice-making energy efficiency model is used to characterize the mapping relationship between ice-making power and ice-making speed under different ambient temperatures, and is called by the optimization control module in step S4.

[0107] The ice storage capacity detection module 103 is used to execute step S3, which collects real-time data on the ice storage capacity in the ice vending machine to obtain the current ice storage capacity. The ice storage capacity detection module 103 may include an ultrasonic level gauge, a weighing sensor, an infrared ranging sensor, an image recognition device, or other sensors capable of estimating the ice storage capacity.

[0108] The optimization control module 104 is used to execute step S4, with the goal of minimizing the total energy consumption in the future control time domain. It takes the ice demand prediction curve, the current ice storage amount and the ice making energy efficiency model as input, and performs rolling optimization on the ice making power sequence under the preset ice storage amount constraint to obtain the optimal ice making power sequence that meets the ice demand requirement and has the lowest total energy consumption.

[0109] The execution drive module 105 is used to execute step S5, extract the ice-making power command corresponding to the current control cycle from the optimal ice-making power sequence, and convert the ice-making power command into the compressor's variable frequency drive signal to continuously adjust the compressor's operating frequency, thereby controlling the ice maker's operation.

[0110] The aforementioned demand forecasting module 101, energy efficiency model storage module 102, ice storage detection module 103, optimization control module 104, and execution drive module 105 can be integrated into the local controller of the ice vending machine. Depending on deployment requirements, the demand forecasting module 101 can also be located in a cloud server, while the energy efficiency model storage module 102, ice storage detection module 103, optimization control module 104, and execution drive module 105 can be located in the local controller of the ice vending machine. The cloud server sends the ice demand forecast curve to the local controller via a wireless communication module. The local controller performs rolling optimization based on the current ice storage and the ice-making energy efficiency model, and executes the ice-making power command for the current control cycle.

[0111] The above modules can be implemented using hardware circuits, software programs, firmware programs, or a combination of hardware and software. Since this system and the aforementioned method are based on the same inventive concept, the specific working process of each module can be referred to the corresponding steps in the aforementioned method embodiments, and will not be repeated here.

[0112] This invention incorporates the nonlinear energy efficiency characteristics of the ice maker into a rolling optimization control process, enabling the ice vending machine to adaptively generate an ice-making power sequence based on future ice demand, current ice storage status, ambient temperature changes, and the equipment's energy efficiency model. Compared to methods that only control start-stop based on current ice storage, switch between high and low power levels based solely on demand forecasts, or simply control peak and off-peak periods, this invention proactively avoids high-power, low-efficiency operating ranges while meeting ice demand and ice storage safety constraints. It prioritizes utilizing the optimal or relatively optimal target power range for advance ice storage, thereby achieving an operating path with lower total energy consumption or lower total electricity costs. This reduces unnecessary energy consumption and improves the operational economy and energy-saving effect of the ice vending machine.

[0113] It will be apparent to those skilled in the art that this disclosure is not limited to the details of the exemplary embodiments described above, and that this disclosure can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of this disclosure is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within this disclosure. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0114] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. An adaptive energy-saving control method for an ice vending machine, characterized in that, Includes the following steps: S1. Obtain historical sales data, future ambient temperature prediction data, and date attribute information from the ice vending machine. Input the historical sales data, ambient temperature prediction data, and date attribute information into the trained demand prediction model to obtain the ice demand prediction curve in the future control time domain. S2. Obtain the ice-making energy efficiency model of the ice vending machine. The ice-making energy efficiency model is a model of the mapping relationship between ice-making power and ice-making speed under different ambient temperatures. Under a fixed ambient temperature, the unit ice production energy consumption of the ice maker exhibits non-linear characteristics as the ice-making power changes, and there exists a target power range that optimizes energy efficiency. S3. Collect the ice storage data in the ice storage compartment of the ice vending machine in real time to obtain the current ice storage volume; S4. With the goal of minimizing the total energy consumption in the future control time domain, the ice demand prediction curve, the current ice storage amount, and the ice-making energy efficiency model are used as inputs. Under the preset ice storage amount constraint, the ice-making power sequence is optimized by rolling optimization to obtain an optimal ice-making power sequence that meets the ice demand requirement and has the lowest total energy consumption. S5. Extract the ice-making power command corresponding to the current control cycle from the optimal ice-making power sequence, and control the ice maker to operate according to the ice-making power command.

2. The adaptive energy-saving control method for ice vending machines according to claim 1, characterized in that, The ice demand forecast curve consists of a set of ice demand data for each discrete future time period, arranged in chronological order.

3. The adaptive energy-saving control method for ice vending machines according to claim 1, characterized in that, The ice-making energy efficiency model described in step S2 is obtained by calibrating the ice-making speed of the ice maker under different ambient temperatures and different operating power, and establishing a two-dimensional mapping relationship with ambient temperature and ice-making power as inputs and ice-making speed as output.

4. The adaptive energy-saving control method for ice vending machines according to claim 1, characterized in that, Step S4, which involves performing a rolling optimization solution for the ice-making power sequence, specifically includes: Using the current ice storage as the initial value of the state variable, multiple candidate ice-making power sequences are generated; For each candidate ice-making power sequence, based on the ice-making energy efficiency model and the ambient temperature of the corresponding time period, the predicted ice-making speed for each control cycle is determined, and combined with the ice demand prediction curve, the predicted ice storage amount at the end of each control cycle in the future control time domain is iteratively calculated to form the ice storage amount change trajectory. Among all candidate ice-making power sequences that satisfy the ice storage constraint, the candidate ice-making power sequence that minimizes the total predicted energy consumption is selected as the optimal ice-making power sequence. The ice storage constraint includes that the ice storage does not exceed the upper limit of the ice storage capacity and is not lower than the lower limit of the safety stock.

5. The adaptive energy-saving control method for ice vending machines according to claim 4, characterized in that, When calculating the total predicted energy consumption, a time-of-use electricity price factor is introduced to make the optimal ice-making power sequence the ice-making power sequence with the lowest total predicted electricity cost.

6. The adaptive energy-saving control method for an ice vending machine according to claim 1, characterized in that, In step S5, the ice maker is controlled to operate according to the ice-making power command. Specifically, the ice-making power command is converted into a variable frequency drive signal for the compressor to continuously adjust the operating frequency of the compressor.

7. The adaptive energy-saving control method for an ice vending machine according to any one of claims 1, 4, or 6, characterized in that, The method executes steps S1 to S5 in a fixed control cycle. Each time it is executed, step S4 outputs only the optimal ice-making power sequence in the future control time domain, while step S5 executes only the first ice-making power command in the optimal ice-making power sequence. In the next control cycle, the entire optimization process is repeated based on the re-acquired current ice storage and the newly acquired ice demand prediction curve.

8. The adaptive energy-saving control method for ice vending machines according to claim 1, characterized in that, The mapping relationship represented by the ice-making energy efficiency model enables the optimal ice-making power sequence to proactively avoid high-power, low-efficiency operating ranges and prioritize the use of the most energy-efficient power ranges for advance ice storage, thereby achieving the lowest total energy consumption while meeting ice demand.

9. An adaptive energy-saving control system for an ice vending machine, characterized in that, include: The demand forecasting module is used to acquire historical sales data, future ambient temperature forecast data, and date attribute information of the ice vending machine. The historical sales data, ambient temperature forecast data, and date attribute information are input into the trained demand forecasting model to obtain the ice demand forecast curve in the future control time domain. An energy efficiency model storage module is used to store the ice-making energy efficiency model of the ice vending machine. The ice-making energy efficiency model is a model of the mapping relationship between ice-making power and ice-making speed under different ambient temperatures. Under a fixed ambient temperature, the unit ice production energy consumption of the ice maker exhibits non-linear characteristics as the ice-making power changes, and there exists a target power range that optimizes energy efficiency. The ice storage volume detection module is used to collect data on the ice storage volume in the ice storage compartment of the ice vending machine in real time to obtain the current ice storage volume. The optimization control module is connected to the demand forecasting module, the energy efficiency model storage module, and the ice storage detection module, respectively. It is used to minimize the total energy consumption in the future control time domain. Taking the ice demand forecasting curve, the current ice storage, and the ice making energy efficiency model as input, it performs rolling optimization on the ice making power sequence under the preset ice storage constraint to obtain an optimal ice making power sequence that meets the ice demand requirement and has the lowest total energy consumption. The execution drive module is connected to the optimization control module and the ice maker of the ice vending machine. It is used to extract the ice-making power command corresponding to the current control cycle from the optimal ice-making power sequence and control the ice maker to operate according to the ice-making power command.

10. The adaptive energy-saving control system for an ice vending machine according to claim 9, characterized in that, The optimization control module is specifically used for: Using the current ice storage as the initial value of the state variable, multiple candidate ice-making power sequences are generated; For each candidate ice-making power sequence, based on the ice-making energy efficiency model and the ambient temperature of the corresponding time period, the predicted ice-making speed for each control cycle is determined, and combined with the ice demand prediction curve, the predicted ice storage amount at the end of each control cycle in the future control time domain is iteratively calculated to form the ice storage amount change trajectory. Among all candidate ice-making power sequences that satisfy the ice storage capacity constraint, the candidate ice-making power sequence with the minimum total predicted energy consumption is selected as the optimal ice-making power sequence. The ice storage capacity constraint includes that the ice storage capacity does not exceed the upper limit of the ice storage capacity and is not lower than the lower limit of the safety stock.

Images