An ice maker energy consumption analysis and intelligent operation optimization management system

Abstract

The application relates to the technical field of commercial energy management data processing, and discloses an ice maker energy consumption analysis and intelligent operation optimization management system, which comprises the following steps: a service end constructs a dynamic operation decision matrix containing binary start-stop instructions based on historical consumption and rate data and issues the matrix; a device end collects inventory level data and calculates a numerical value gradient in a sliding window, and determines an index offset according to gradient-offset mapping logic; the device end generates virtual inventory level data by subtracting the index offset from the inventory level data, and searches start-stop instruction data in the matrix according to a time slice index to serve as scheduling basis; and the application realizes adaptive response to sudden business demands without changing a static strategy matrix by means of data dimension reduction and index key reconstruction mechanisms based on consumption rate characteristics, and solves the hysteresis problem of an offline planning model in a dynamic environment.

Description

Technical Field

[0001] This invention relates to an energy consumption analysis and intelligent operation optimization management system for ice makers, belonging to the field of commercial energy management data processing technology. Background Technology

[0002] Current commercial refrigeration equipment operation and management relies on a physical state feedback mechanism for immediate response. By monitoring the liquid level or temperature parameters of the ice storage tank in real time, the refrigeration compressor is triggered to supplement cooling capacity when the monitored value reaches a preset threshold. This type of control logic, which meets current physical demand and ensures basic business supply, reveals limitations in data processing logic in commercial scenarios involving multi-dimensional resource constraints. Existing technologies attempt to introduce intelligent monitoring methods to improve management efficiency, but still have limitations in the economic and adaptive dimensions of control strategies. For example, the Chinese invention patent with authorization announcement number CN113701413B discloses an ice-making system for household ice makers based on artificial intelligence analysis. Although this solution assesses the ice-making level and provides feedback on progress by monitoring temperature and cooling rate, it is essentially a passive state monitoring and information interaction. The control logic does not incorporate the energy value factor of time-of-use electricity pricing into the decision-making closed loop, and cannot achieve dynamic optimization of operating costs. When abnormal performance indicators are detected, only warning signals are generated, lacking the ability to automatically compensate for equipment aging or environmental disturbances, making it difficult to meet the stringent requirements of unattended operation, cost control, and business continuity in commercial scenarios.

[0003] With the widespread adoption of time-of-use pricing policies, the shortcomings of traditional control methods have become apparent. Existing technologies lack the ability to calculate the time dimension of energy value, leading to equipment frequently running cooling commands during peak electricity price periods, resulting in increased operating costs. The hard-coupled mechanism of real-time physical feedback ignores the cold storage buffer properties of ice as a phase change medium and fails to utilize this property to achieve staggered production and consumption configurations over time. Faced with pulse-like demand disturbances in commercial scenarios, the system, lacking an inventory buffer strategy, faces the risk of rapid depletion of physical inventory when encountering sudden high-frequency ice extraction, leading to business interruptions. Constrained by the computing power and cost of edge devices, existing solutions struggle to deploy complex predictive models or global optimization algorithms at the terminal. Equipment cannot dynamically adjust control strategies based on energy efficiency degradation characteristics, and long-term operation can easily lead to physical performance drift, resulting in supply and demand imbalances and control failures.

[0004] Therefore, how to construct a data processing method with low computing power requirements, utilizing time-sharing energy value, and having adaptive resource scheduling capabilities has become the technical problem to be solved by this invention. Summary of the Invention

[0005] To address the problems mentioned in the background art, the technical solution of the present invention is as follows: An ice maker energy consumption analysis and intelligent operation optimization management system, comprising a data processing server for performing strategy calculations and a device execution terminal for performing on-site data interaction:

[0006] The data processing server is used to construct and issue a dynamic operation decision matrix based on historical business consumption data and time-based rate data. The dynamic operation decision matrix is ​​a two-dimensional data structure with the discrete time slice index as the first dimension and the discrete inventory level index as the second dimension. The matrix element values ​​are binary start and stop instruction data.

[0007] The device execution end is used to acquire inventory level data representing the current resource inventory, calculate the numerical descent gradient of the inventory level data within a preset sliding window, and determine the index offset corresponding to the numerical descent gradient based on a preset gradient-offset mapping logic. The device execution end performs an arithmetic operation on the inventory level data by subtracting the index offset to generate virtual inventory level data, and maps the current system clock to a time slice index. Using the time slice index and virtual inventory level data as retrieval keys, the device execution end retrieves the corresponding binary start / stop command data from the locally stored dynamic operation decision matrix, and outputs this binary start / stop command data as the basis for switching resource production states. The gradient-offset mapping logic is limited to the following: as the numerical descent gradient increases, the index offset monotonically and non-decreasing, enabling the device execution end to retrieve the start command data corresponding to the low inventory level coordinates in the dynamic operation decision matrix through input state data reconstruction.

[0008] Preferably, the device execution end includes a gradient data processing module; the gradient data processing module is used to maintain a circular data queue containing the most recently preset number of sampled values ​​in memory to realize data interception of a preset sliding window; the gradient data processing module reads the inventory level data at the current moment and the inventory level data at the historical moment in the circular data queue in each data processing cycle, and generates instantaneous gradient data representing the current business resource consumption rate by calculating the difference between the two and dividing by the time interval value.

[0009] Preferably, the device execution end includes a virtual index generation unit; the virtual index generation unit stores gradient-offset mapping logic, which is manifested as a non-linear data lookup table; the non-linear data lookup table sets at least two gradient grading threshold data, when the numerical descent gradient is in the first numerical interval, the output is an index offset of zero, when the numerical descent gradient is in the second numerical interval higher than the first numerical interval, the output is an index offset of the first fixed step size, and when the numerical descent gradient is in the third numerical interval higher than the second numerical interval, the output is an index offset of the second fixed step size greater than the first fixed step size; when the virtual index generation unit generates virtual inventory level data, if the calculation result is negative, it executes data clamping logic to set the result to zero.

[0010] Preferably, the dynamic operation decision matrix includes a standard decision data table generated based on the rated energy efficiency model and a conservative decision data table generated based on the decay energy efficiency model; the distribution density of start-up command data for low-rate periods in the standard decision data table is lower than the distribution density of start-up command data for the same period in the conservative decision data table; the device execution end is configured with data source addressing switching logic, which is used to switch the read pointer between the storage address of the standard decision data table and the storage address of the conservative decision data table according to the performance index data representing the health of the device.

[0011] Preferably, the device execution end also includes a status data discretization unit; the status data discretization unit is used to receive continuously changing resource inventory signals and convert them into discretized inventory level integers according to a preset quantization step size; the value range of the discretized inventory level integers is limited to be consistent with the address index range of the second dimension of the dynamic operation decision matrix; when the device execution end performs a retrieval operation, it directly uses the virtual inventory level data as the address offset of the matrix row or column to perform a single data reading operation.

[0012] Preferably, the device execution end also includes a performance index inversion module; the performance index inversion module is used to record the actual time consumed in the process of resource inventory recovering from a preset low threshold to a preset high threshold. The performance index inversion module is used to obtain the baseline theoretical time data corresponding to the current ambient temperature data. And calculate the performance index data according to the following formula. : Among them, the baseline theoretical time consumption data The baseline parameters are pre-set on the equipment execution end based on the output efficiency model under standard operating conditions for the data processing server.

[0013] Preferably, the data source addressing switching logic is configured as follows: at the end of each resource recovery cycle, the calculated performance index data is compared with the preset drift threshold data; when the performance index data is less than the drift threshold data, the read pointer is positioned to the standard decision data table; when the performance index data is greater than or equal to the drift threshold data, the read pointer is positioned to the conservative decision data table, and a maintenance work order data package containing abnormal feature identifiers is generated at the same time.

[0014] Preferably, the dynamic operation decision matrix has an asymmetric hysteresis data distribution characteristic; the asymmetric hysteresis data distribution characteristic means that in the data distribution of the matrix, for the same time slice index, the inventory level critical data for switching from the stop instruction value to the start instruction value is smaller than the inventory level critical data for switching from the start instruction value to the stop instruction value; the difference between the two critical data dynamically increases as the time-sharing rate value of the corresponding time slice increases, so as to construct a digital filtering dead zone for high-frequency instruction switching at the data level.

[0015] Preferably, the data processing server includes a strategy optimization calculation engine; the strategy optimization calculation engine is used to extract time-series features from the acquired historical business consumption data to generate a typical daily consumption model, and combines time-of-use rate data and equipment energy efficiency data to use a dynamic programming algorithm to solve for the state sequence with the lowest total cost under the stockout rate constraint; the strategy optimization calculation engine binarizes the state sequence and maps it into a two-dimensional data structure, and updates the data structure to the local database of the device execution end through a network transmission protocol.

[0016] Preferably, the device execution end also includes data circuit breaker protection logic; the data circuit breaker protection logic is used to monitor the numerical deviation between virtual inventory level data and inventory level data in real time; when the numerical deviation continuously exceeds the preset safety boundary value, the data circuit breaker protection logic temporarily suspends the retrieval operation for the dynamic operation decision matrix and directly sets the start instruction data until the inventory level data is restored to the preset safety threshold data.

[0017] Compared with the prior art, the beneficial effects of the present invention are:

[0018] 1. In the energy consumption analysis of ice makers, an asymmetric data processing architecture is constructed for cloud-based strategy generation and terminal table lookup execution. This resolves the contradiction between complex cost optimization models and limited computing resources of edge devices. The system pre-processes the multi-dimensional optimization calculation based on historical consumption characteristics, time-of-use rate data, and equipment energy efficiency curves on the server side, and the results are solidified into a two-dimensional dynamic operation decision matrix containing definite start and stop instructions. The terminal device obtains control instructions only based on the current time and inventory level index matrix, transforming the high-order linear programming problem into a deterministic data addressing operation processing mechanism. This reduces the computational load and hardware cost at the edge, eliminates the risk of control failure due to real-time calculation delays or algorithm non-convergence, and ensures the accurate and stable implementation of the optimal operation strategy based on business rules in a low-computing-power environment.

[0019] 2. A virtual inventory mapping mechanism based on inventory consumption gradients is introduced to improve the system's ability to adapt to impulsive demand disturbances without changing the static decision matrix structure. By monitoring the rate of inventory level decline within a sliding time window in real time, the corresponding coordinate offset is calculated, and the inventory status data of the input decision matrix is ​​logically reconstructed. The reconstruction process utilizes the negative correlation between inventory level and start-up probability in the decision matrix. When rapid consumption characteristics are detected, the system is automatically guided to retrieve high-priority production instructions. Through logical transformation at the data preprocessing level, the inherent defect of static scheduling strategies being unable to respond to sudden surges in business traffic is solved, ensuring energy cost control targets while achieving nonlinear protection for business supply continuity.

[0020] 3. By utilizing the statistical characteristics of operational data, an implicit feedback loop for equipment health is constructed to enable automatic adaptation of control strategies to the degradation of equipment physical performance. By comparing the actual operating time of a complete work cycle with the theoretical baseline time, the equipment efficiency deviation ratio is quantified. Based on the ratio, the query pointer is dynamically switched between a preset standard decision matrix and a conservative decision matrix. Based on the time residual analysis strategy addressing and adjustment mechanism, the system can detect performance drift caused by condenser blockage or component aging without adding physical detection components, and automatically degrade to a high-protection operation mode. This maintains supply and demand matching business availability throughout the entire life cycle of equipment physical condition deterioration, avoiding supply chain disruptions caused by the disconnect between digital models and physical entities. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the system data processing architecture for the dynamic decision matrix of this invention.

[0022] Figure 2 This is a comparison diagram of the equipment's peak-shifting operation status and strategy under the time-of-use pricing environment of this invention;

[0023] Figure 3 This is a flowchart illustrating the dynamic demand response and energy efficiency adaptive maintenance logic of the present invention. Detailed Implementation

[0024] This specific embodiment provides an ice maker energy consumption analysis and intelligent operation optimization management system. The following embodiments are only used to explain the present invention and are not intended to limit the present invention.

[0025] This invention provides an ice maker energy consumption analysis and intelligent operation optimization management system, which consists of a data processing server that performs strategy calculations and an equipment execution terminal that performs on-site data interaction. The two establish a bidirectional data communication link through TCP / IP protocol or MQTT IoT protocol. The system achieves asymmetric decoupling of computing load by deploying high-dimensional global cost optimization calculation tasks on the data processing server and deploying deterministic data indexing and instruction execution tasks on the equipment execution terminal. In specific deployment, the data processing server runs on a cloud server cluster and integrates a historical data storage unit, a rate database, and a strategy optimization calculation engine. The equipment execution terminal is embedded in the microcontroller unit (MCU) of the commercial ice maker and includes a non-volatile memory for storing strategy data, a sensor interface for collecting inventory status signals, and an I / O output port for controlling the compressor relay.

[0026] To address the multi-dimensional constraints between time-of-use electricity pricing and inventory safety in commercial scenarios, this system employs a discretized control mechanism based on a two-dimensional decision matrix. The data processing server acquires historical energy consumption data of the target ice maker and time-of-use tariff data for its region on a daily basis. The strategy optimization calculation engine discretizes the continuous time axis into time slice indices with fixed step sizes. The continuous ice storage tank level signal is discretized into a fixed gradient inventory level index. The server-side algorithm, based on dynamic programming or Monte Carlo simulation, uses minimizing operating costs as the objective function and ensuring that inventory does not fall below a safety threshold as a constraint. It then reverse-engineers the optimal start-stop sequence for the next 24 hours. The calculation result is mapped and solidified into a two-dimensional data structure, namely the dynamic operation decision matrix. Each element in this matrix has a value of either 0 or 1, where 1 represents a start command and 0 represents a stop command. After this data structure is sent to the device execution end, the control logic on the device end does not involve floating-point operations or predictive inference, but instead performs extremely low-latency coordinate indexing operations, that is, based on the current clock... and current inventory Direct Read The value is used to drive the relay. This mechanism ensures that the device can still maintain the optimal operating strategy based on business rules according to the locally stored matrix when the network is disconnected.

[0027] To address the issue of sudden, impulsive demand in commercial operations causing physical inventory to be rapidly depleted, the device execution end is configured with virtual inventory mapping logic based on consumption gradients. This allows for real-time responses to dynamic disturbances without modifying the static matrix data distributed from the cloud. The device execution end maintains a circular buffer in memory to store the inventory level sample values ​​within the most recent preset time window. In each data processing cycle, the gradient data processing module reads the current inventory level. Inventory level at the start of the window Calculate the difference between the two and normalize it to obtain the inventory decline gradient. The system has a pre-defined non-linear gradient-offset mapping table that defines the gradient value and the index offset. The correspondence between them, when a decrease in inventory gradient is detected. When the surge threshold is exceeded, the mapping table outputs a non-zero positive integer offset. The device execution end performs arithmetic operations. Generate virtual inventory levels that are numerically lower than the actual physical inventory. The device execution end utilizes this virtual inventory level Instead of using the actual inventory level as the index key, the dynamic operation decision matrix is ​​queried, thereby retrieving the activation command corresponding to the low inventory area in the matrix. This process is achieved by actively shifting the input coordinates.

[0028] To address the issue of actual energy efficiency deviating from the theoretical model due to the aging of physical components in ice makers, the system constructs a closed-loop feedback and multi-modal strategy switching mechanism based on time residuals. At least two sets of decision matrices are pre-generated and distributed in the cloud, namely, a standard decision matrix generated based on rated energy efficiency. and conservative decision matrix generated based on energy decay efficiency In the conservative decision matrix, for the same time and inventory coordinates, the density of start-up instructions with a bit set to 1 is higher than that in the standard decision matrix, in order to reserve greater capacity redundancy. The equipment execution end uses the efficiency index inversion module to monitor in real time the actual running time of the ice maker from low liquid level to full capacity signal trigger. And extract the theoretical reference duration corresponding to the current ambient temperature from the memory. The module is based on the formula Calculate the efficiency deviation ratio The ratio As a dynamic pointer control variable, when When the drift is less than the preset drift threshold, the data read pointer points to the standard decision matrix. The storage address; when When the value is greater than or equal to this threshold, the pointer automatically switches to point to the conservative decision matrix. The storage address ensures that the control system automatically adapts to the physical state when the device's ice-producing efficiency decreases due to condenser blockage or refrigerant micro-leakage.

[0029] Example 1: In the actual operation scenario of a chain tea beverage store, the ice maker energy consumption analysis and intelligent operation optimization management system faces a complex operating condition with concurrent high-rate periods and sudden pulse business demands. At this time, the system clock... The electricity price is in the peak range, and the current inventory level of the physical ice storage is... Located within a pre-defined downtime energy-saving zone in the cloud, based on a dynamic operation decision matrix that only considers static cost optimization. The binarized start / stop command data corresponding to this coordinate point is zero, indicating that the equipment should remain off to avoid expensive power consumption. However, due to a surge in ice-collecting frequency caused by a sudden increase in group orders, the gradient data processing module embedded in the equipment's execution end calculates the numerical gradient of the inventory level data within the current sliding window by reading the circular data queue. If the value exceeds the preset first value range, and the system strictly follows the static matrix cost control logic, the physical inventory will be depleted before the next control cycle, leading to business interruption. To resolve this dynamic supply and demand imbalance, the device execution end does not rely on real-time intervention from the cloud, but instead triggers local gradient-offset mapping logic, decreasing the gradient based on the larger value. Determine the non-zero index offset The device execution end performs arithmetic operations. This generates virtual inventory level data that is numerically lower than the current actual physical inventory. The device execution end then uses this virtual inventory level data. Replace actual inventory level As the second dimension of the search key value.

[0030] Dynamically running decision matrix stored locally During relocation, since the matrix was built with embedded safety constraints for enabling low inventory levels, the virtual coordinates are automatically indexed to the storage area where the enabling instruction data is 1. This closes the compressor relay without modifying the static strategy matrix data structure or increasing the edge-end prediction computing load. Simultaneously, considering the physical condition of reduced heat exchange efficiency due to dust accumulation in the condenser, the performance index inversion module compares the actual time consumption data at the end of the ice-making cycle. Compared with the baseline theoretical time data Calculate the performance index data If the data drift threshold is exceeded, the data source addressing switching logic will respond immediately, automatically switching the read pointer of the next cycle from the storage address of the standard decision data table to the storage address of the conservative decision data table. The conservative decision data table has a higher density of start instruction distribution under the same time slice index, thereby compensating for the decline in physical capacity by increasing the cooling time in subsequent operation. This process solves the engineering problem that static planning models cannot adapt to high-frequency business disturbances and equipment physical performance drift at extremely low computing power cost through input state data reconstruction and dynamic switching of address pointers.

[0031] Example 2: This example uses commercial refrigeration equipment and simulated dynamic workload to construct a rigorous control test platform. It consists of two identical commercial air-cooled ice makers (rated ice production 120kg / 24h, rated power 850W). The test environment simulates the daily operating conditions of a standard chain tea shop, including periodic ice-removing loads and random pulse loads. During the test, a high-precision electricity meter (accuracy 0.5S level) was used to record system power consumption, and a liquid level sensor and electronic weighing system were used to monitor ice inventory and production in real time. The test introduced random environmental temperature disturbances (25°C). ±5 ) and simulated grid voltage fluctuations (220V±10%).

[0032] To verify the cost optimization capability of this solution at the static planning level, a comparative sample group and sample group 1 of this invention were set up. The comparative sample group adopted the traditional physical liquid level trigger control logic, that is, when the inventory is lower than 80%, the refrigeration is started immediately until the inventory is full. Sample group 1 of this invention fully deployed a control strategy based on a dynamic operation decision matrix. The matrix was constructed based on the local real time-of-use electricity price data (peak electricity price 1.2 yuan / kWh, off-peak electricity price 0.3 yuan / kWh), and the safety inventory threshold was set at 30%. The test was run for 72 hours. The data showed that the cumulative running time of the comparative sample group during the peak electricity price period accounted for 42%, resulting in a higher average unit ice-making cost. In contrast, sample group 1 of this invention, through the constraint of matrix instructions, shifted a large number of refrigeration tasks to the off-peak electricity price period at night, and its peak operation rate was reduced to 12%. See Table 1. The data shows that sample group 1 of this invention reduces operating costs while ensuring that the total ice production meets the demand.

[0033] Table 1: Comparison of Energy Consumption and Cost Data under Different Control Strategies

[0034]

[0035] To verify the dynamic response capability of the virtual inventory mapping logic based on consumption gradient in dealing with sudden surges in demand, sample group 2 of this invention was set up. In the 36th hour of the experiment, a high-frequency ice extraction operation lasting 30 minutes was simulated, with the consumption rate reaching 3 times the normal rate. At this time, it was during the peak electricity price period, and the static matrix indicated shutdown. The gradient data processing module of sample group 2 of this invention calculated the inventory decline gradient in real time. If the surge threshold is exceeded, the system then determines the index offset based on the gradient-offset mapping table. And generate virtual inventory levels. This virtual inventory level The system was successfully induced to retrieve the start command in the matrix. As a control, if this mapping logic were removed and only the static matrix was run, the system would suffer a physical inventory breakdown due to rigid adherence to cost rules. The experimental results confirmed that after introducing the virtual inventory mapping mechanism, the system's minimum physical inventory under sudden loads consistently remained above the safety line, verifying the mechanism's crucial role in ensuring business continuity. To verify the effectiveness of the time residual-based energy efficiency drift adaptive mechanism, sample group 3 of this invention and an out-of-range control group were set up. In the experiment, the energy efficiency degradation caused by equipment aging was simulated by artificially blocking the condenser heat sink, causing the actual ice production efficiency to decrease by approximately 20%. Sample group 3 of this invention calculated the efficiency deviation ratio. Detected If the drift exceeds the preset threshold (15%), it will automatically switch to a conservative decision matrix. This matrix compensates for the production capacity gap by increasing the cooling time, ensuring that the daily ice production meets the standard. In contrast, the out-of-range control group set the drift threshold to an excessively high 50% (exceeding the upper limit of the reasonable range), which caused the system to fail to switch strategies in time when energy efficiency decreased, ultimately resulting in a shortage caused by insufficient ice production.

[0036] Example 3: This example combines Figures 1 to 3 This document describes an energy consumption analysis and intelligent operation optimization management system for an ice maker, such as... Figure 1 As shown, historical business consumption data on the server side, after time-series feature extraction and time-of-use rate data processed by the peak-valley electricity price model, is fed into the strategy optimization calculation engine. Dynamic programming is used to solve the minimum cost state sequence to generate a dynamic operation decision matrix containing standard decision data tables and conservative decision data tables, which is then distributed synchronously over the network. On the equipment side, inventory level data, after real-time liquid level acquisition, enters the gradient data processing module to calculate the numerical descent gradient. The virtual index generation unit converts this gradient into a virtual inventory level based on the gradient offset mapping logic. The matrix retrieval unit uses the time slice index mapped by the system clock and the virtual inventory level as keys to retrieve binarized start / stop command data locally and outputs addressing control signals. Simultaneously, the performance index inversion module compares the actual consumption time with the baseline consumption time to calculate the deviation ratio. The deviation ratio The driver's data source addressing switching logic switches between standard and conservative decision data table pointers.

[0037] like Figure 2 As shown, the horizontal axis represents time, the left vertical axis represents the electricity price (yuan / kWh), and the right vertical axis represents the operating status (0 out, 1 in). The time-of-use electricity price curve in the figure shows obvious peaks and troughs. The traditional strategy's operating status curve shows that the equipment follows physical demand and maintains operating status 1 during the high electricity price periods of 10:00-12:00 and 18:00-20:00. However, the operating status curve of this invention shows that the equipment, through a pre-planned strategy, concentrates the main cooling periods between 06:00-08:00 and after 22:00 during the low electricity price periods, and maintains an outage status 0 during the high-rate intervals. Figure 3 As shown, external data sources are used to construct dynamic operational decision matrices and support the extraction of time-series features. Physical ice-making equipment operates based on retrieved binary start / stop commands. On one hand, the system generates virtual inventory levels by calculating numerical descent gradients and performing gradient offset mapping, which serves as input for command retrieval. On the other hand, the system calculates performance index data in real time. Logical judgment shows when When the threshold is reached, the switch to standard / conservative decision data table is triggered, and a maintenance work order data package is generated and delivered to the operation and maintenance management personnel.

[0038] Example 4: In large-scale cold chain logistics distribution centers, ice machine clusters not only need to cope with dynamic business loads but also face more severe challenges from power grid fluctuations and ambient temperature changes. To address this complex scenario, this example constructs an adaptive fault-tolerant control mechanism based on full lifecycle performance monitoring. In the first phase of the experiment, the operating conditions of the ice machines in the cold chain center under high-temperature conditions in summer are simulated, where the ambient temperature... It has remained at 35 The above exceeds the equipment's standard rated operating conditions (25). If the system relies solely on a static energy efficiency model, the actual ice production efficiency will be lower than expected, resulting in insufficient ice-making windows during off-peak electricity pricing periods to fill the ice storage. To address this model mismatch black box, the device's execution end utilizes built-in temperature sensors to collect data in real time. It also calls upon the locally stored temperature-energy efficiency compensation curve, obtains it through offline calibration experiments, and clarifies the theoretical ice production efficiency correction coefficient under different ambient temperatures. Before the start of each control cycle, the policy optimization calculation engine reads the current... Query the corresponding correction coefficient And utilize the corrected effective ice production rate The system recalculates the required ice-making time. Data shows that after introducing this temperature compensation logic, even under extreme high temperatures, the system can automatically extend the operating time during off-peak hours to ensure that the inventory is always at full capacity the next morning.

[0039] In the second phase of the experiment, the heat dissipation efficiency of an ice maker's condenser was simulated by gradually accumulating dust. This was achieved by artificially covering the condenser surface with dustproof nets of varying densities, thus altering the actual heat transfer coefficient of the equipment. Gradually decreasing to 70% of the rated value, during which time the performance index inversion module continuously monitors the actual runtime of each complete ice-making cycle. When the efficiency deviation ratio is monitored for three consecutive cycles If all values ​​exceed the preset first-level warning threshold (10%) but are below the switching threshold (15%), the system does not immediately switch matrices. Instead, it triggers soft correction logic, adjusting the current standard decision matrix. Based on this, a secondary verification is performed on all 0 instructions (shutdown). If the inventory level at that moment is in the critical zone and future forecast demand is high, the instruction is temporarily modified to 1 (start) to slightly increase capacity reserves. However, as the dust accumulation deepens further, leading to... When the switching threshold of 15% is finally exceeded, the data source addressing switching logic decisively intervenes, switching the control pointer from... Permanently switch to conservative decision matrix The matrix incorporates a more aggressive cooling strategy, sacrificing some economic efficiency in exchange for guaranteed production capacity.

[0040] Finally, regarding the impact of grid voltage fluctuations on compressor energy efficiency, this embodiment introduces voltage adaptive protection logic. Under the simulated grid voltage drop to 198V (90% of the rated voltage), the compressor motor slip increases, leading to a decrease in the cooling energy efficiency ratio (COP). At this time, the device actuator detects the abnormal input voltage through the voltage sensor and immediately activates the peak avoidance delay strategy. If the current situation is not an emergency shortage (inventory is greater than the safety threshold), the system will temporarily suspend the current start command and enter the standby monitoring state until the voltage returns to normal or the inventory reaches the lower limit. This avoids the compressor from doing ineffective work in the low voltage and low energy efficiency range, which not only protects the motor coils but also further reduces the overall energy consumption of the system. This embodiment constructs a closed-loop adaptive control system by introducing three targeted repair modules: ambient temperature compensation, multi-level energy efficiency drift correction, and voltage adaptive protection.

[0041] Example 5: This example constructs a standardized offline calibration and data filling procedure, aiming to obtain the real performance characteristics of the ice maker under different operating conditions through controlled physical experiments, and to construct a high-precision temperature-energy efficiency compensation curve and drift threshold benchmark accordingly; a full-temperature-range energy efficiency calibration experiment is performed by placing the target ice maker in a constant temperature and humidity test chamber and setting the ambient temperature. From 10 Up to 45 With 5 A gradient was applied to the step size, and at each temperature point, the ice maker was run continuously for three complete ice-making cycles, recording the actual ice production in each cycle. With power consumption Calculate the actual energy efficiency ratio at this temperature. , each test point Compared with standard rated operating condition 25 Energy efficiency ratio Normalization is performed to obtain the correction coefficient. By using a polynomial fitting method, discrete test points are fitted into a continuous correction curve, and the curve data is discretized and stored in the non-volatile memory of the device execution end as the data source of the temperature-energy efficiency compensation curve.

[0042] Aging simulation and drift threshold definition experiments were conducted. Under standard operating conditions, different degrees of dust accumulation and blockage were simulated by artificially installing dust filters of different mesh sizes at the condenser air inlet, and the blockage rate was defined. Increasing from 0% to 50%, the duration of a single ice-making cycle for the ice maker was recorded at each clogging rate. Experimental data shows that when When the critical value of 30% is reached, A rapid, non-linear increase occurred, and ice production began to fall below the design limit. Based on this inflection point data, it was determined that the system would switch to a conservative decision matrix. The optimal entry point is to calculate the efficiency deviation ratio corresponding to that inflection point. and will The drift threshold is set as the data source addressing switching logic. This procedure ensures that the system can trigger policy switching based on physical measured boundaries rather than empirical guesses, thereby enabling a response to the aging state of the equipment.

[0043] Example 6: This example provides a standardized pre-deployment calibration and model building procedure to ensure the stability and control accuracy of the system under different business environments and individual equipment differences. This procedure is designed to be a mandatory procedure after the system's first power-on or after major maintenance. Through automated data acquisition and parameter optimization, the core control model is localized and adapted. Static parameter self-checks and inventory quantification calibration are performed. After the system starts, the ice maker is controlled to enter full-power ice-making mode until the full-capacity sensor signal is triggered. During this process, the equipment execution end continuously monitors and records the real-time power curve of the compressor and the high and low pressure data of the refrigerant circuit to construct the current benchmark energy efficiency fingerprint of the equipment. A controlled ice-melting test is performed, the compressor is turned off and the ice-melting valve is opened, and the mass flow rate of the ice-melting water is recorded in real time using a high-precision electronic scale. By establishing a mapping relationship between ice-melting time and ice-melting mass, the system reverse-calibrates the nonlinear correspondence function between the analog reading of the ice storage tank level sensor and the actual ice inventory. It is solidified into non-volatile memory and used as the physical basis for subsequent discretization of inventory levels.

[0044] Dynamic load response testing and gradient threshold optimization were performed. The system simulated three different ice-collecting load modes with varying gradients: steady consumption mode, rapid consumption mode, and rapid pulse mode. Under each mode, the real-time decline curve of inventory level data was recorded, and the corresponding instantaneous decline gradient was calculated. Through statistical analysis, the optimal gradient boundary point for distinguishing between normal and abnormal pulse consumption was determined. This boundary point was then used as the gradient grading threshold data in the gradient-offset mapping table. The minimum gradient value in the rapid pulse mode was set as the threshold for triggering the first-level offset. The maximum gradient value under rapid consumption mode is set as the reset threshold to ensure sensitive response to sudden demand while avoiding false triggers caused by fluctuations in regular ice production. Finally, the online energy efficiency drift monitoring model is initialized. After completing the above calibration, the system enters a 72-hour trial operation and learning period. During this period, the efficiency index inversion module runs silently in the background, collecting actual time data for no less than 20 complete ice-making cycles. With the corresponding ambient temperature Using these data points, the system employs the least squares method to fit the baseline theoretical time consumption curve of the current device. This curve will replace the factory-preset universal model and serve as the efficiency deviation ratio for subsequent calculations. Personalized benchmarks.

[0045] Example 7: During the system initialization offline calibration phase, the gradient grading threshold data is derived from the statistical analysis of the target device's historical inventory level data for 30 working days. Gradient samples from all inventory decline periods are extracted, and the cumulative probability distribution function is calculated. The 90th percentile of the distribution function is established as the trigger threshold for the first-level offset gradient. The 98th percentile was established as the threshold for triggering the second-level offset gradient. Corresponding index offset The physical meaning is determined by the time slice length. It calculates the theoretical maximum ice production under rated operating conditions within a single time slice and converts it into the corresponding discrete inventory level step size. The first fixed step size is set to Round up to the nearest whole number, and set the second fixed step size as... Rounding up to the nearest whole number, this calibration procedure ensures virtual inventory levels. When inputting the matrix, the system retrieves instructions for the device to start ahead of time, compensating for the future capacity gap caused by the pulse demand.

[0046] In the dynamic decision matrix generation mechanism, a deterministic decay coefficient is introduced when the strategy optimization calculation engine constructs the conservative decision data table. The boundary conditions are reconstructed, and a dynamic programming algorithm is used to solve the minimum-cost state sequence process. The standard decision data table is based on the rated ice production rate of the equipment. The state transition equation's capacity constraint parameters are used in the conservative decision data table. Re-solve for capacity constraint parameters, attenuation coefficient The energy efficiency degradation rate is determined by measured values ​​when the condenser blockage rate reaches a critical value of 30%. Based on a secondary solution mechanism for degradation parameters, the conservative decision data table mathematically outputs a wider range of start-up command sequences under the same inventory and time coordinates. This physically offsets the actual cooling capacity degradation of the equipment by increasing the cooling time; baseline theoretical time consumption data. The following standardized lookup procedure is executed: The equipment execution terminal pre-stores a mapping table relating ambient temperature to ice-making time. This mapping table is then processed by a constant temperature environment laboratory to obtain the ambient temperature... From 10 Up to 45 With 5 The step-by-step test yielded the average physical time taken for the equipment to go from empty to full at each temperature point as a baseline value. During real-time operation, the performance index inversion module reads the current ambient temperature sensor value and uses linear interpolation to calculate the corresponding theoretical reference time from the mapping table. Participation in performance indicator data Calculation, ensure It only reflects performance deviations caused by aging or dirt blockage of the equipment's own components, excluding background interference caused by ambient temperature fluctuations.

[0047] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0048] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. An ice maker energy consumption analysis and intelligent operation optimization management system, characterized in that, This includes a data processing server for performing strategy calculations and a device execution end for performing on-site data interaction: The data processing server is used to construct and issue a dynamic operation decision matrix based on historical business consumption data and time-based rate data. The dynamic operation decision matrix is ​​a two-dimensional data structure with the discrete time slice index as the first dimension and the discrete inventory level index as the second dimension. The matrix element values ​​are binary start and stop instruction data. The device execution end is used to acquire inventory level data representing the current resource inventory, calculate the numerical descent gradient of the inventory level data within a preset sliding window, and determine the index offset corresponding to the numerical descent gradient based on a preset gradient-offset mapping logic. The device execution end performs an arithmetic operation on the inventory level data by subtracting the index offset to generate virtual inventory level data, and maps the current system clock to a time slice index. Using the time slice index and virtual inventory level data as retrieval keys, the device execution end retrieves the corresponding binary start / stop command data from the locally stored dynamic operation decision matrix, and outputs this binary start / stop command data as the basis for switching resource production states. The gradient-offset mapping logic is limited to the following: as the numerical descent gradient increases, the index offset monotonically and non-decreasing, enabling the device execution end to retrieve the start command data corresponding to the low inventory level coordinates in the dynamic operation decision matrix through input state data reconstruction.

2. The ice maker energy consumption analysis and intelligent operation optimization management system according to claim 1, characterized in that, The device execution end includes a gradient data processing module; the gradient data processing module is used to maintain a circular data queue containing the most recently preset number of sampled values ​​in memory. In each data processing cycle, the gradient data processing module reads the inventory level data at the current moment and the inventory level data at the historical moment in the circular data queue. By calculating the difference between the two and dividing it by the time interval value, instantaneous gradient data representing the current business resource consumption rate is generated.

3. The ice maker energy consumption analysis and intelligent operation optimization management system of claim 1, wherein, The device execution end includes a virtual index generation unit; the virtual index generation unit stores gradient-offset mapping logic, which is manifested as a non-linear data lookup table; the non-linear data lookup table sets at least two gradient grading threshold data, when the numerical gradient is in the first numerical interval, the output is an index offset of zero, when the numerical gradient is in the second numerical interval higher than the first numerical interval, the output is an index offset of the first fixed step size, and when the numerical gradient is in the third numerical interval higher than the second numerical interval, the output is an index offset of the second fixed step size greater than the first fixed step size; when the virtual index generation unit generates virtual inventory level data, if the calculation result is negative, it executes data clamping logic to set the result to zero.

4. The ice maker energy consumption analysis and intelligent operation optimization management system of claim 1, wherein, The dynamic operation decision matrix includes a standard decision data table generated based on the rated energy efficiency model and a conservative decision data table generated based on the decay energy efficiency model. The distribution density of start-up command data in the standard decision data table for low-rate periods is lower than that of the conservative decision data table for the same period. The device execution end is configured with data source addressing switching logic, which is used to switch the read pointer between the storage address of the standard decision data table and the storage address of the conservative decision data table based on the performance index data that characterizes the health of the device.

5. The ice maker energy consumption analysis and intelligent operation optimization management system according to claim 4, characterized in that, The device execution end also includes a status data discretization unit; the status data discretization unit is used to receive continuously changing resource inventory signals and convert them into discretized inventory level integers according to a preset quantization step size; The range of values ​​for the discrete inventory level integers is limited to the range of the address index of the second dimension of the dynamic operation decision matrix; when the device execution end performs a retrieval operation, it uses the virtual inventory level data as the address offset of the matrix row or column for a single data read operation.

6. The ice maker energy consumption analysis and intelligent operation optimization management system of claim 4, wherein, The device execution end further comprises an efficiency index inversion module; the efficiency index inversion module is configured to record actual time consumption data of the recovery of the resource stock from the preset low threshold to the preset high threshold ; The performance index inversion module is used to obtain the baseline theoretical time data corresponding to the current ambient temperature data. And calculate the performance index data according to the following formula. : Among them, the baseline theoretical time consumption data The baseline parameters are pre-set on the equipment execution end based on the output efficiency model under standard operating conditions for the data processing server.

7. The ice maker energy consumption analysis and intelligent operation optimization management system of claim 6, wherein, The data source addressing switching logic is configured as follows: at the end of each resource recovery cycle, the calculated performance index data is compared with the preset drift threshold data; when the performance index data is less than the drift threshold data, the read pointer is positioned to the standard decision data table; when the performance index data is greater than or equal to the drift threshold data, the read pointer is positioned to the conservative decision data table, and a maintenance work order data package containing abnormal feature identifiers is generated at the same time.

8. The ice maker energy consumption analysis and intelligent operation optimization management system of claim 1, wherein, The dynamic operation decision matrix has an asymmetric hysteresis data distribution characteristic. The asymmetric hysteresis data distribution characteristic means that in the data distribution of the matrix, for the same time slice index, the inventory level critical data for switching from the stop instruction value to the start instruction value is smaller than the inventory level critical data for switching from the start instruction value to the stop instruction value. The difference between the two critical data increases dynamically as the time-sharing rate value of the corresponding time slice increases, so as to construct a digital filtering dead zone for high-frequency instruction switching at the data level.

9. The ice maker energy consumption analysis and intelligent operation optimization management system of claim 1, wherein, The data processing server includes a strategy optimization calculation engine. This engine extracts time-series features from historical business consumption data to generate a typical daily consumption model. It then combines time-of-use rate data and equipment energy efficiency data with a dynamic programming algorithm to solve for the state sequence that minimizes total cost while meeting the stockout rate constraint. The engine binarizes the state sequence and maps it into a two-dimensional data structure. This data structure is then updated to the local database on the device execution side via a network transmission protocol.

10. The ice maker energy consumption analysis and intelligent operation optimization management system of claim 1, wherein, The device execution end also includes data circuit breaker protection logic; the data circuit breaker protection logic is used to monitor the numerical deviation between virtual inventory level data and inventory level data in real time; when the numerical deviation exceeds the preset safety boundary value continuously, the data circuit breaker protection logic temporarily suspends the retrieval operation for the dynamic operation decision matrix and directly sets the start instruction data until the inventory level data is restored to the preset safety threshold data.

Images