Intelligent vending machine system with self-adaptive multi-area temperature control function

Abstract

The invention belongs to the technical field of intelligent temperature control, and discloses an intelligent vending machine system with a self-adaptive multi-area temperature control function. Comprising the steps that multi-mode heat flux data in the vending machine and on the surfaces of commodities are collected through a sensor array; iterative fitting is carried out on the collected data, and a three-dimensional thermal field distribution diagram is constructed; based on isotherm density and heat flow vector direction analysis, the interior of the vending machine is divided into independent thermal control units, and a thermal plume intersection path is recognized; generating a multi-source energy dynamic distribution matrix by combining the target temperature with the energy demand index and the thermal plume interference weight of the real-time thermal field distribution; carrying out convection path adjustment on the thermal control unit with the risk of heat overflow; the energy demand index is updated in real time by continuously monitoring the microenvironment temperature fluctuation frequency, accurate multi-area temperature control is achieved, and the energy utilization efficiency and the commodity fresh-keeping quality are greatly improved.

Description

Technical Field

[0001] This invention relates to the field of intelligent temperature control technology, and more specifically, to an intelligent vending machine system with adaptive multi-zone temperature control function. Background Technology

[0002] Multi-zone temperature control in retail automation is a key element in improving the operational efficiency of smart vending machines, especially in scenarios where food, beverages, and temperature-sensitive goods are sold together, where it has significant economic value. Traditional vending machine temperature control systems mainly employ single-zone designs or simple mechanical partitions. However, these methods are not only energy-inefficient but also fail to meet the differentiated temperature requirements of various products, resulting in a product loss rate as high as 15-20%. With the development of control technology, temperature control methods such as variable frequency refrigeration and zoned temperature control have been applied. However, these systems generally suffer from severe thermal interference between temperature zones, slow response, and insufficient accuracy.

[0003] Thermodynamics and fluid dynamics, as fundamental theories of physics, have found widespread application in the field of microenvironment temperature control in recent years. The conduction, convection, and radiation of heat energy in enclosed spaces follow specific laws; through precise analysis of the thermal field distribution and dynamic characteristics of heat flow, accurate control of microenvironment temperature can be achieved. Studies have shown that heat exchange processes in multi-temperature zones are extremely complex, especially in thermal boundary regions, where airflow and temperature difference effects lead to the convergence of thermal plumes and the formation of temperature gradient zones. While subtle, these phenomena can significantly reduce temperature control accuracy and waste energy.

[0004] Existing technologies struggle to effectively capture and manage the dynamic changes in the thermal field within the complex environment of vending machines. These changes involve multiple physical quantities and typically manifest as nonlinear fluctuations in spatial temperature distribution, making reliable monitoring difficult with traditional single-point or sparse sensor networks. During vending machine operation, factors such as product replenishment, customer pickup, and ambient temperature fluctuations cause drastic changes in the internal thermal field, especially after the door is opened, where rapid exchange of hot and cold air leads to over-response of the temperature control system. Traditional temperature control systems employ fixed-parameter PID control or simple ON / OFF strategies, lacking the ability to analyze the thermal field topology in real time and failing to identify and predict the path of heat plumes and their impact on different temperature zones. Inside vending machines with high-density product arrangement, the heat flow dynamics are even more complex, with the heat exchange effects between different products compounding, further increasing the difficulty of temperature control. Existing systems also lack adaptive energy allocation mechanisms and thermal risk warning functions, failing to optimize energy utilization efficiency while ensuring temperature control accuracy, resulting in high operating costs and an inability to meet the precise preservation requirements of different products.

[0005] In view of this, the present invention proposes an intelligent vending machine system with adaptive multi-zone temperature control function to solve the above problems. Summary of the Invention

[0006] To overcome the aforementioned deficiencies of the prior art and to achieve the above objectives, the present invention provides the following technical solution:

[0007] A smart vending machine system with adaptive multi-zone temperature control includes:

[0008] The data acquisition module is used to collect data from the inside of the vending machine and the surface of the goods based on a preset sensor array, and to generate multimodal heat flux data.

[0009] The map construction module is used to iteratively fit the real-time acquired multimodal heat flux data to obtain a three-dimensional thermal field distribution map.

[0010] The region calibration module is used to perform real-time analysis based on the isothermal linear density and heat flow vector direction within the three-dimensional thermal field distribution map. Based on the real-time analysis results, the internal space of the vending machine is divided into several independent thermal control units. Virtual thermal boundaries are defined for adjacent independent thermal control units, and the confluence paths of heat plumes between adjacent independent thermal control units are identified based on the virtual thermal boundaries.

[0011] The response generation module is used to obtain the energy demand index between the target temperature of each independent thermal control unit and the instantaneous three-dimensional thermal field distribution map, and combine the interference weight of the thermal plume confluence path to generate a multi-source energy dynamic allocation matrix.

[0012] The response control module is used to identify the heat overflow risk of independent thermal control units based on the multi-source energy distribution matrix, and to adjust the convection path of independent thermal control units with heat overflow risk.

[0013] The response feedback module is used to continuously monitor the internal space of the vending machine after adjustment, obtain the corresponding micro-environment temperature fluctuation frequency, and update the corresponding energy demand index in real time based on the micro-environment temperature fluctuation frequency.

[0014] Furthermore, the process of generating the three-dimensional thermal field distribution map includes:

[0015] The collected multimodal heat flux data is spatially continuous to obtain the continuous temperature field distribution corresponding to the internal space of the vending machine, and the temperature gradient vector of each location in the internal space of the vending machine is obtained based on it.

[0016] The heat transfer coefficients between products in the microenvironment are obtained based on a pre-built reference database of the thermal properties of the products, and the radiative heat flux data between the corresponding products in the microenvironment are obtained based on the heat transfer coefficients and temperature gradient vectors.

[0017] The obtained radiative heat flux data is spatially fused with the continuous temperature field distribution, and the thermal field is reconstructed to obtain a three-dimensional thermal field distribution map.

[0018] Furthermore, the process of spatially continuum processing of the collected multimodal heat flux data includes:

[0019] The collection location of contact temperature data in the vending machine's internal space is obtained from the corresponding multimodal heat flux data, and spatial gridding is performed on it to obtain the inner wall temperature distribution feature vector, which includes the temperature gradient change rate and the boundary temperature conduction coefficient.

[0020] The temperature feature vector of the commodity surface is obtained based on infrared radiation data in multimodal heat flux data. The temperature feature vector includes surface temperature uniformity and hot spot distribution index.

[0021] Data fusion and spatial interpolation are performed based on the obtained inner wall temperature distribution feature vector and the product surface temperature feature vector to obtain the continuous temperature field distribution corresponding to the vending machine content space.

[0022] Furthermore, the process of obtaining the heat transfer coefficient includes:

[0023] Based on the thermal property parameters and surface area of ​​the commodity materials stored in the pre-built commodity thermal property reference database, the convective heat transfer coefficient between commodities is obtained; and the radiative heat transfer coefficient is obtained simultaneously; and a comprehensive heat transfer coefficient is obtained based on the convective heat transfer coefficient and the radiative heat transfer coefficient.

[0024] Furthermore, the process of obtaining the convergence path of thermal plumes includes:

[0025] Isothermal surfaces are extracted from the three-dimensional thermal field distribution map to obtain isothermal surfaces at different temperature levels; and isothermal linear density indices are obtained based on the curvature characteristics and density distribution of the isothermal surfaces.

[0026] Based on the temperature gradient vector in the three-dimensional thermal field distribution map, the divergence and curl of the heat flux vector direction are obtained, and the characteristic parameters of the heat flux vector are obtained.

[0027] The coupling analysis of isothermal linear density index and heat flux vector characteristic parameters identifies spatial regions with similar heat transfer behavior, and based on this, the internal space of the vending machine is divided into several independent thermal control units.

[0028] A virtual thermal boundary network is formed by connecting the temperature gradient abrupt change points at the boundaries of adjacent independent thermal control units; and by analyzing the temperature gradient discontinuity in the virtual thermal boundary network, the thermal plume convergence path between independent thermal control units is obtained.

[0029] Furthermore, the process of obtaining the multi-source energy dynamic allocation matrix includes:

[0030] Set the target temperature range for each independent heat control unit, and determine the temperature control accuracy based on the preservation requirements of the goods in the area corresponding to each independent heat control unit;

[0031] Based on the instantaneous three-dimensional thermal field distribution map, the deviation vector between the actual temperature distribution and the target temperature in each independent thermal control unit is obtained, and the energy change required for each independent thermal control unit to reach the target temperature is obtained based on it, thus obtaining the energy demand index.

[0032] Flow analysis is performed on the thermal plume confluence path to obtain the heat transfer capacity of each path and the degree of impact on adjacent units, and the interference weight corresponding to the thermal plume confluence path is obtained based on it.

[0033] Based on the energy demand index and corresponding disturbance weight of each independent thermal control unit, a multi-objective optimization problem is constructed and solved to obtain a multi-source energy dynamic allocation matrix.

[0034] Furthermore, the process of obtaining the interference weights includes:

[0035] Multiple virtual detection planes are set up on the identified thermal plume confluence path, and time-series sampling is performed on the corresponding virtual detection planes to record the temperature distribution and velocity vector of each virtual detection plane at different time points. Based on this, a dynamic flow characteristic curve is constructed, which includes the heat flux change rate and directional stability index.

[0036] The total heat transfer of the thermal plume confluence path per unit time is obtained based on the constructed dynamic flow characteristic curve, and the heat distribution density and velocity gradient of the path cross section are analyzed; combined with the geometry and transmission direction of the thermal plume confluence path, the heat transfer index of each thermal plume confluence path is obtained.

[0037] Temperature fluctuation spectrum analysis is performed on the virtual thermal boundary corresponding to adjacent independent thermal control units to identify the temperature disturbance mode of the thermal plume on the adjacent independent thermal control units. Based on this, the ratio of the temperature fluctuation amplitude caused by the thermal plume to the self-stabilization capability of the adjacent units is calculated to obtain the influence degree coefficient matrix. The influence degree coefficient matrix characterizes the interference intensity of the thermal plume on the temperature field stability of each adjacent unit.

[0038] By combining the heat transfer capacity index and the influence coefficient matrix, the interference weight of each thermal plume convergence path in the overall temperature control process is calculated.

[0039] Furthermore, the process of adjusting the convection path includes:

[0040] Heat overflow risk is identified based on a multi-source energy dynamic allocation matrix, resulting in independent thermal control units with heat overflow risk. Topological analysis is performed on the convection channels between multiple independent thermal control units to identify key convection nodes between each independent thermal control unit. Based on this, the convection path of the independent thermal control units with heat overflow risk is reconstructed, and the heat distribution is verified.

[0041] Furthermore, the process of continuously monitoring the internal space of the adjusted vending machine to obtain the corresponding micro-environment temperature fluctuation frequency, and updating the corresponding energy demand index in real time based on the micro-environment temperature fluctuation frequency, includes:

[0042] Continuous multimodal data acquisition is performed on the internal space of the adjusted vending machine, and the temperature time series corresponding to each independent thermal control unit is obtained based on the data.

[0043] Frequency domain analysis was performed on the temperature time series of each independent thermal control unit to obtain the main frequency component and fluctuation amplitude of temperature fluctuations, and to obtain the frequency characteristics of microenvironment temperature fluctuations.

[0044] The frequency characteristics of temperature fluctuations are compared and analyzed with preset stability thresholds, and the temperature control effect and stability level of each independent thermal control unit are evaluated based on these thresholds to obtain a stability score. The corresponding energy demand index is then updated in real time based on the obtained stability score.

[0045] Furthermore, the process of identifying heat spillover risks based on a multi-source energy dynamic allocation matrix includes:

[0046] The energy density of each independent thermal control unit is obtained and compared with a preset density threshold. Areas exceeding the threshold are marked as potential heat overflow areas.

[0047] The temperature gradient deviation between the potential heat overflow area and the adjacent independent thermal control unit is evaluated. When the temperature gradient deviation exceeds a preset deviation threshold and the duration exceeds a preset duration, it is finally marked as an independent thermal control unit with a risk of heat overflow.

[0048] The technical effects and advantages of the intelligent vending machine system with adaptive multi-zone temperature control function of this invention are as follows:

[0049] This invention enables precise capture and management of the complex thermal field distribution and dynamic changes within vending machines. It transforms the previously unpredictable convergence paths of heat plumes into controllable temperature regulation bases. This changes the temperature control mode of smart vending machines, shifting from passive, single-zone temperature control to active, multi-zone precise adjustment, significantly reducing energy consumption and product spoilage rates. In practical applications, vending machine operators no longer need to provide excessive cooling energy for the entire machine space; instead, they can precisely allocate energy based on real-time thermal field analysis results, improving the system's energy efficiency. The adaptive thermal control characteristics of this invention fully adapt to dynamic changes in products and environmental fluctuations, avoiding excessive temperature fluctuations and inter-zone thermal interference problems in traditional methods, while effectively extending the shelf life and freshness of temperature-sensitive products. Attached Figure Description

[0050] Figure 1 This is a schematic diagram of the intelligent vending machine system with adaptive multi-zone temperature control function according to the present invention. Detailed Implementation

[0051] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0052] Example 1

[0053] Please see Figure 1 As shown, the intelligent vending machine system with adaptive multi-zone temperature control function described in this embodiment includes:

[0054] The data acquisition module is used to collect data from the inside of the vending machine and the surface of the goods based on a preset sensor array, and to generate multimodal heat flux data.

[0055] The map construction module is used to iteratively fit the real-time acquired multimodal heat flux data to obtain a three-dimensional thermal field distribution map.

[0056] The region calibration module is used to perform real-time analysis of the isothermal linear density and heat flux vector direction within the three-dimensional thermal field distribution map. Based on the real-time analysis results, the internal space of the vending machine is divided into several independent thermal control units. Virtual thermal boundaries are defined for adjacent independent thermal control units, and the confluence paths of heat plumes between adjacent independent thermal control units are identified based on the virtual thermal boundaries.

[0057] The response generation module is used to obtain the energy demand index between the target temperature of each independent thermal control unit and the instantaneous three-dimensional thermal field distribution map, and combine the interference weight of the thermal plume confluence path to generate a multi-source energy dynamic allocation matrix.

[0058] The response control module is used to identify the heat overflow risk of independent thermal control units based on the multi-source energy distribution matrix, and to adjust the convection path of independent thermal control units with heat overflow risk.

[0059] The response feedback module is used to continuously monitor the internal space of the vending machine after adjustment, obtain the corresponding micro-environment temperature fluctuation frequency, and update the corresponding energy demand index in real time based on the micro-environment temperature fluctuation frequency.

[0060] It should be further explained that, in the specific implementation process, the acquisition of multimodal heat flux data includes:

[0061] Real-time temperature data streams within the vending machine's interior are collected using a pre-deployed temperature sensor array to obtain contact temperature data.

[0062] The infrared sensor array pre-deployed inside the vending machine is used to scan the surface of the goods inside the vending machine with infrared radiation to obtain the infrared radiation intensity matrix of the temperature distribution of the goods surface. The infrared radiation intensity matrix is ​​then calibrated for emissivity to obtain infrared radiation data.

[0063] The obtained contact temperature data and infrared radiation data are spatiotemporally synchronized and registered, and combined with the spatial geometric information of the vending machine's internal space to form multimodal heat flux data.

[0064] In this embodiment, firstly, PT1000 platinum resistance temperature sensors are deployed at key nodes within the vending machine's interior space to form an encircling temperature sensing layout, with each sensor having a unique digital identifier and three-dimensional spatial coordinates. Then, based on these sensors, the temperature data stream within the corresponding vending machine's interior space is collected in real time, and the data from each sensor is uploaded and aggregated via a CAN bus to obtain raw temperature data. The collected raw temperature data is then filtered using wavelet transform to remove electromagnetic interference and transient noise, and a high-precision timestamp (accurate to milliseconds) and corresponding three-dimensional spatial coordinate information are added to each temperature data point, forming contact-type temperature data containing a time-space-temperature triplet. The key nodes include the various partitions, side walls, top, and bottom within the vending machine.

[0065] Simultaneously, an array of infrared thermal imaging sensors (such as the FLIRLEPT series infrared sensor matrix) is installed at the center of the top of the vending machine. The sensor has infrared detection capability in the 8-14μm wavelength range and can simultaneously acquire infrared radiation intensity information of multiple product surfaces. The sensor scanning angle and focal length are set to ensure complete coverage of all product shelves inside the vending machine. Then, based on the set sensor scanning angle and focal length, the corresponding product surfaces are infrared scanned to obtain the infrared radiation intensity distribution of each product surface at different times, forming a two-dimensional infrared radiation intensity matrix. According to the emissivity characteristics of different product packaging materials (e.g., metal packaging emissivity 0.1-0.3, plastic packaging emissivity 0.9-0.95), the two-dimensional infrared radiation intensity matrix is ​​calibrated to eliminate the influence of the vending machine's internal ambient temperature on infrared measurement and convert the infrared radiation intensity into the true surface temperature value, forming infrared radiation data. The emissivity calibration refers to subtracting the infrared radiation intensity reflected by the corresponding product surface from the collected infrared radiation intensity.

[0066] Then, a standard Cartesian coordinate system is established with the bottom left front corner of the vending machine's internal space as the origin. The measurement data from the temperature sensor and the infrared sensor are unified into the Cartesian coordinate system, ensuring that the data collected by different types of sensors at the same time have temporal consistency. Combined with the geometric structure information inside the vending machine (such as shelf dimensions, product placement, and ventilation vent locations), a multi-dimensional data structure containing spatial location, temperature value, time information, and sensor type is constructed to form a multimodal heat flux dataset. Finally, the multimodal heat flux data is organized and compressed according to the time series, while maintaining the integrity and recoverability of the data, forming a continuous multimodal heat flux data stream that supports real-time processing and historical backtracking.

[0067] It should be further explained that, in the specific implementation process, the generation process of the three-dimensional thermal field distribution map includes:

[0068] The collected multimodal heat flux data is spatially continuous to obtain the continuous temperature field distribution corresponding to the internal space of the vending machine, and the temperature gradient vector of each location in the internal space of the vending machine is obtained based on it.

[0069] The heat transfer coefficients between products in the microenvironment are obtained based on a pre-built reference database of the thermal properties of the products, and the radiative heat flux data between the corresponding products in the microenvironment are obtained based on the heat transfer coefficients and temperature gradient vectors.

[0070] The obtained radiative heat flux data is spatially fused with the continuous temperature field distribution, and the thermal field is reconstructed to obtain a three-dimensional thermal field distribution map.

[0071] In this embodiment, the collection location of contact temperature data in the vending machine's internal space is obtained from the corresponding multimodal heat flux data, and spatially meshed and mapped to obtain an inner wall temperature distribution feature vector, which includes the temperature gradient change rate and the boundary temperature conduction coefficient; and the temperature feature vector of the product surface is obtained based on the infrared radiation data in the multimodal heat flux data, which includes the surface temperature uniformity and the hot spot distribution index; based on the obtained inner wall temperature distribution feature vector and the product surface temperature feature vector, data fusion and spatial interpolation are performed to obtain the continuous temperature field distribution corresponding to the vending machine's internal space;

[0072] Then, a three-dimensional unsteady heat conduction equation was established based on heat transfer theory. ,in For temperature, For time, Let Q0 be the thermal diffusivity and Q0 be the internal heat source term. To transmit medium density, For specific heat capacity, The partial differential equation is used to numerically solve the constructed heat conduction equation, obtaining the temperature gradient vector at each grid point in three-dimensional space. Where x, y, and z represent the three dimensions of the three-dimensional space, respectively; the temperature gradient vector indicates the direction and intensity of the heat flow, providing a vector basis for subsequent heat flow analysis.

[0073] Next, the convective heat transfer coefficient between products is obtained based on the thermal property parameters and surface area of ​​the product materials stored in the pre-built product thermal property reference database. Where Nu is the Nusselt number, k is the fluid thermal conductivity, and L is the length of the product; and the radiative heat transfer coefficient is obtained simultaneously. ;in, Emissivity is the surface emissivity of a commodity (i.e., emissivity is a measure of the ability of an object's surface to radiate energy outward). Given the Stefan-Boltzmann constant and T as the absolute temperature, the comprehensive heat transfer coefficient is obtained. Among them, the pre-built reference database of thermal properties of commodities contains several thermal property parameters, including specific heat capacity, thermal conductivity, surface roughness and emissivity of various commodities;

[0074] Get the viewing angle factor between any two product surfaces Perspective Factor This represents the proportion of radiant energy emitted from the surface of the i-th product that is received by the surface of the j-th product.

[0075] Among them, perspective factor In the formula, and Let represent the areas of the i-th and j-th items, respectively; and Let represent the angles between the normal to the center point on the surface of the i-th and j-th commodities and the line connecting them; and Let represent the tiny surface elements on the i-th and j-th product surfaces, respectively; r represents the distance between the two surface elements; where the connecting line represents the actual propagation path of the radiated energy;

[0076] Furthermore, based on the obtained perspective factors Obtain the radiant heat flow between the corresponding products In the formula, and Let represent the absolute temperatures of the i-th and j-th product surfaces, respectively;

[0077] Based on the obtained radiative heat flux and combined with the geometry and spatial layout of the goods, a microenvironment radiative heat flux model is established. This model can obtain the radiative heat flux distribution in a small-scale space around the goods, and improve the accuracy of radiative heat transfer acquisition by considering multiple reflections and scattering effects, thus obtaining detailed microenvironment radiative heat flux data between goods.

[0078] Finally, the continuous temperature field distribution data and radiative heat flux data are registered and superimposed in three-dimensional space. Using a finite element mesh as a unified spatial discretization framework, temperature values ​​and heat flux vector information are stored simultaneously for each mesh cell in the finite element mesh. Combined with heat transfer boundary conditions (such as adiabatic boundary, isothermal boundary, and convective boundary) and the distribution of internal heat sources, a three-dimensional thermal field distribution map containing temperature distribution and heat flux density vector is generated. This map not only shows the temperature distribution inside the vending machine, but also reveals the path and intensity of heat transfer, providing a scientific thermodynamic basis for the formulation of temperature control strategies.

[0079] It should be further explained that, in the specific implementation process, the process of spatially continuum processing of the collected multimodal heat flux data includes:

[0080] The collection location of contact temperature data in the vending machine's internal space is obtained from the corresponding multimodal heat flux data, and spatial gridding is performed on it to obtain the inner wall temperature distribution feature vector, which includes the temperature gradient change rate and the boundary temperature conduction coefficient.

[0081] The temperature feature vector of the commodity surface is obtained based on infrared radiation data in multimodal heat flux data. The temperature feature vector includes surface temperature uniformity and hot spot distribution index.

[0082] Data fusion and spatial interpolation are performed based on the obtained inner wall temperature distribution feature vector and the product surface temperature feature vector to obtain the continuous temperature field distribution corresponding to the vending machine content space.

[0083] In this embodiment, the process of spatially continuum processing of the collected multimodal heat flux data specifically includes the following steps:

[0084] First, the contact temperature data from the corresponding multimodal heat flux data is collected at the locations within the vending machine's interior space, and then spatially meshed to obtain the inner wall temperature distribution feature vector. Specifically, the collection locations corresponding to the contact temperature data are calibrated in three dimensions. The vending machine's interior space is divided into grid cells with adaptive scales, where the cell size is automatically adjusted according to temperature gradient changes, with increased grid density in areas of rapid temperature change and decreased density in areas of less rapid change. Gaussian filtering is applied to the time-series temperature data at each collection location to eliminate random noise interference. The temperature gradient between adjacent collection locations is calculated, and the temperature change rate is calculated using a time window (60 seconds). Based on heat conduction theory, combined with the vending machine's inner wall material properties and thickness parameters, the boundary temperature conduction coefficient λ = q / (A·ΔT / d) is calculated, where q is the heat flux, A is the heat transfer area, ΔT is the temperature difference, and d is the material thickness. Finally, the inner wall temperature distribution feature vector is formed.

[0085] Secondly, the temperature feature vector of the commodity surface is obtained based on the infrared radiation data in the multimodal heat flux data. Specifically, a commodity surface temperature distribution matrix is ​​constructed based on the collected infrared radiation data. The root mean square deviation and average temperature of each commodity surface are calculated, and the surface temperature uniformity is obtained based on them. The surface temperature uniformity = 1 - (root mean square deviation ÷ average temperature), with a value range of [0,1]. The closer the value is to 1, the more uniform the temperature distribution. Hot spots on the commodity surface are identified, a hot spot temperature threshold is set, the area ratio and average temperature of the hot spot area are calculated, and a hot spot distribution index is defined based on them. The hot spot distribution index = area ratio of hot spot area × (average temperature of hot spot area - average temperature of commodity surface) ÷ average temperature of commodity surface to quantify the degree of commodity surface temperature anomaly. The surface temperature uniformity and the hot spot distribution index are combined to form the commodity surface temperature feature vector.

[0086] Finally, based on the obtained inner wall temperature distribution feature vector and the product surface temperature feature vector, data fusion and spatial interpolation are performed to obtain the continuous temperature field distribution corresponding to the vending machine's internal space. Specifically, a three-dimensional spatial discrete point set is constructed, including inner wall sampling points and product surface detection points, forming an initial temperature field spatial skeleton. The discrete point set is initially interpolated based on the thin plate spline interpolation method (TPSI), and the interpolation results are corrected based on the heat conduction equation.

[0087] A Bayesian fusion method is used to integrate the feature vectors of the inner wall and the product surface, and spatial adaptive variational interpolation is applied to the data fusion results to ensure that the temperature field meets the energy conservation constraint. A multi-resolution grid optimization algorithm is used to refine the grid in key areas to improve accuracy and reduce computational resource consumption. Through iterative calculation (usually 5 to 10 iterations), convergence is determined when the maximum temperature difference between two adjacent iterations is less than 0.1°C, resulting in the final continuous temperature field distribution. The spatial adaptive variational interpolation is an intelligent spatial data reconstruction method that achieves smooth transitions between data points by minimizing the global energy function. This method includes key steps such as adaptive grid density adjustment, dynamic weight allocation, and iterative solution, which can achieve high-precision spatial interpolation while maintaining physical constraints. The adaptive grid density adjustment mechanism automatically optimizes the computational grid based on local data distribution characteristics and gradient changes. A coarser grid is used in sparse data areas to save computational resources, while a finer grid is used in dense data areas or areas with drastic gradient changes to improve accuracy, ensuring that the interpolation results satisfy both global consistency and accurately depict local details.

[0088] It should be further explained that, in the specific implementation process, the process of obtaining the thermal plume convergence path includes:

[0089] Isothermal surfaces are extracted from the three-dimensional thermal field distribution map to obtain isothermal surfaces at different temperature levels; and isothermal linear density indices are obtained based on the curvature characteristics and density distribution of the isothermal surfaces.

[0090] Based on the temperature gradient vector in the three-dimensional thermal field distribution map, the divergence and curl of the heat flux vector direction are obtained, and the characteristic parameters of the heat flux vector are obtained.

[0091] The coupling analysis of isothermal linear density index and heat flux vector characteristic parameters identifies spatial regions with similar heat transfer behavior, and based on this, the internal space of the vending machine is divided into several independent thermal control units.

[0092] A virtual thermal boundary network is formed by connecting the temperature gradient abrupt change points at the boundaries of adjacent independent thermal control units; and by analyzing the temperature gradient discontinuity in the virtual thermal boundary network, the thermal plume convergence path between independent thermal control units is obtained.

[0093] In this embodiment, multiple temperature thresholds are set (e.g., an isothermal surface is set every 0.5℃), and three-dimensional isothermal surfaces corresponding to the three-dimensional thermal field distribution map are extracted based on them to form a layered set of isothermal surfaces. Each isothermal surface represents a surface composed of points with equal temperature in space. Geometric analysis is performed on the extracted isothermal surfaces to obtain the principal curvature and average curvature of each isothermal surface. Areas with large curvature indicate drastic temperature changes and may be key areas where heat is concentrated or dispersed.

[0094] The number of isothermal surfaces per unit volume is counted, and an isothermal linear density index is defined based on this. ;in This refers to the number of isothermal surfaces within a local area. For the volume of a local region, high-density regions represent steep temperature gradients and are potential candidate regions for thermal boundaries.

[0095] Then, the temperature gradient vector in the three-dimensional thermal field distribution map is obtained, and vector differentiation is performed on the gradient vector field to obtain the divergence. In the formula, The sum of the second-order partial derivatives of the temperature gradient vector in all dimensions is represented; the divergence represents the convergence or divergence of heat flow, with positive values ​​indicating heat flow convergence (heat source) and negative values ​​indicating heat flow divergence (cold source); and the curl corresponding to the temperature gradient vector is obtained simultaneously. Although the temperature gradient field is usually a conservative field (with zero curl), local curl may exist under complex geometry and boundary conditions, which can be used to identify the vortex structure of heat flow. Furthermore, the divergence and curl information are integrated into heat flow vector characteristic parameters, including the singularity type (source point, sink point, saddle point) and intensity of the vector field.

[0096] Next, the isothermal linear density index and heat flux vector characteristic parameters are coupled and analyzed to find their correlation, and a coupling strength index is defined based on this correlation. ,in , , The weighting coefficients are determined through expert knowledge and experience; regions with coupling strength exceeding a preset coupling threshold are identified and used as critical boundary regions.

[0097] Based on the spatial distribution and connectivity of the critical boundary region, the internal space of the vending machine is divided into multiple connected sub-regions. The thermal field characteristics within each sub-region are relatively uniform, and there are obvious thermal boundaries between different sub-regions. Each sub-region is defined as an independent thermal control unit, and each unit is assigned a unique identifier and attribute information (such as volume, product type, target temperature range, etc.).

[0098] Based on the obtained temperature gradient vector, the abrupt change points existing in the boundary region between adjacent independent thermal control units are obtained, and adjacent abrupt change points are connected to form a continuous boundary surface; and a virtual thermal boundary network is constructed based on it, which defines the boundary line between different independent thermal control units; the abrupt change point refers to the point in the boundary of adjacent independent thermal control units where the temperature gradient vector deviation exceeds a preset deviation threshold.

[0099] Finally, based on the virtual thermal boundary network, boundary heat flux analysis is performed on the divided independent thermal control units to obtain the heat flux density vectors on the boundary surfaces between each unit. (Where k is thermal conductivity), the direction and magnitude of the heat flow vector are analyzed to identify the dominant heat flow direction and locate the interface regions where the heat flow intensity is greater than a preset intensity threshold. These regions are the main channels for cross-unit heat transfer. Based on the principles of data acquisition fluid dynamics (CFD), the natural and forced convection phenomena inside the vending machine are analyzed to identify the thermal plumes formed by rising hot air and their development trajectory in space. The movement path of the thermal plumes is tracked through streamline visualization technology, and the intersection points and convergence areas of different thermal plumes are identified. A network topology describing the cross-unit thermal plume transfer is constructed. This network includes the starting point, ending point, intermediate nodes, and connecting paths of the thermal plumes. Each path has corresponding heat flow intensity and influence range parameters, forming a network of thermal plume intersection paths. This region calibration method based on thermodynamic principles can objectively identify the natural thermal zones inside the vending machine, avoiding the subjectivity of manual division and providing a scientific spatial basis for subsequent precise temperature control.

[0100] It should be further explained that, in the specific implementation process, the acquisition process of the multi-source energy dynamic allocation matrix includes:

[0101] Set the target temperature range for each independent heat control unit, and determine the temperature control accuracy based on the preservation requirements of the goods in the area corresponding to each independent heat control unit;

[0102] Based on the instantaneous three-dimensional thermal field distribution map, the deviation vector between the actual temperature distribution and the target temperature in each independent thermal control unit is obtained, and the energy change required for each independent thermal control unit to reach the target temperature is obtained based on it, thus obtaining the energy demand index.

[0103] Flow analysis is performed on the thermal plume confluence path to obtain the heat transfer capacity of each path and the degree of impact on adjacent units, and the interference weight corresponding to the thermal plume confluence path is obtained based on it.

[0104] Based on the energy demand index and corresponding disturbance weight of each independent thermal control unit, a multi-objective optimization problem is constructed and solved to obtain a multi-source energy dynamic allocation matrix.

[0105] In this embodiment, a database of commodity preservation temperature requirements is first constructed, which includes parameters such as the optimal storage temperature range, temperature change tolerance, and preservation time for different types of commodities. Based on the composition of commodities in each independent heat control unit, a weighted average method is used to determine the comprehensive target temperature range of the corresponding independent heat control unit. For units with mixed commodities, the most stringent temperature standard is selected as the control target.

[0106] Based on the instantaneous three-dimensional thermal field distribution map, the actual temperature distribution data within each independent thermal control unit is extracted. And construct a temperature deviation vector based on it. ,in The target temperature is then determined; furthermore, by analyzing the spatial distribution of the temperature deviation vector, overheated and undercooled regions are identified.

[0107] Then, based on the first law of thermodynamics, the heat change required for each independent thermal control unit to reach the target temperature is obtained, and the formula is applied. Where m is the mass of the product within the independent thermal control unit, c1 is the specific heat capacity of the product, and V is the air volume within the unit. air density, The specific heat capacity of air, Q represents the change in air temperature, and Q represents the amount of heat required.

[0108] At the same time, the time constant of heat transfer is obtained. Where h is the heat transfer coefficient, A is the heat transfer area, and the energy demand index is defined. This index takes into account both the required heat and the heat transfer time, reflecting the unit's urgency regarding cooling or heating energy. A higher value indicates a need for faster temperature adjustment.

[0109] Next, multiple virtual detection planes are set up on the identified thermal plume confluence path, and time-series sampling is performed on the corresponding virtual detection planes to record the temperature distribution and velocity vector of each virtual detection plane at different time points. Based on this, a dynamic flow characteristic curve is constructed, which includes the heat flux change rate and directional stability index.

[0110] The total heat transfer of the thermal plume confluence path per unit time is obtained based on the constructed dynamic flow characteristic curve, and the heat distribution density and velocity gradient of the path cross section are analyzed; combined with the geometry and transmission direction of the thermal plume confluence path, the heat transfer index of each thermal plume confluence path is obtained.

[0111] Temperature fluctuation spectrum analysis is performed on the virtual thermal boundary corresponding to adjacent independent thermal control units to identify the temperature disturbance mode of the thermal plume on the adjacent independent thermal control units. Based on this, the ratio of the temperature fluctuation amplitude caused by the thermal plume to the self-stabilization capability of the adjacent units is calculated to obtain the influence degree coefficient matrix. The influence degree coefficient matrix characterizes the interference intensity of the thermal plume on the temperature field stability of each adjacent unit.

[0112] By combining the heat transfer capacity index and the influence coefficient matrix, the interference weight of each thermal plume convergence path in the overall temperature control process is calculated.

[0113] Finally, a multi-objective optimization problem is established based on the obtained interference weights. The objective functions include minimizing total energy consumption, minimizing temperature control deviation, and minimizing system response time. The constraints include power limits for each refrigeration unit, energy balance constraints, and temperature control accuracy requirements. An energy allocation matrix is ​​constructed, where the matrix elements represent the proportion of energy provided by energy sources (such as compressors, evaporators, and fans) to independent thermal control units. A genetic algorithm is applied to solve the problem and obtain the optimal energy allocation scheme. The dynamic update of the matrix elements is based on real-time changes in the energy demand index and interference weights, forming a multi-source dynamic energy allocation matrix. This matrix not only guides the coordinated operation of each refrigeration unit but also considers the impact of cross-unit thermal interference, realizing system-level energy optimization allocation. This energy allocation method based on thermodynamic modeling and optimization theory can maximize energy utilization efficiency while meeting temperature control requirements.

[0114] It should be further explained that, in the specific implementation process, the process of obtaining the interference weight includes:

[0115] Multiple virtual detection planes are set up on the identified thermal plume confluence path, and time-series sampling is performed on the corresponding virtual detection planes to record the temperature distribution and velocity vector of each virtual detection plane at different time points. Based on this, a dynamic flow characteristic curve is constructed, which includes the heat flux change rate and directional stability index.

[0116] The total heat transfer of the thermal plume confluence path per unit time is obtained based on the constructed dynamic flow characteristic curve, and the heat distribution density and velocity gradient of the path cross section are analyzed; combined with the geometry and transmission direction of the thermal plume confluence path, the heat transfer index of each thermal plume confluence path is obtained.

[0117] Temperature fluctuation spectrum analysis is performed on the virtual thermal boundary corresponding to adjacent independent thermal control units to identify the temperature disturbance mode of the thermal plume on the adjacent independent thermal control units. Based on this, the ratio of the temperature fluctuation amplitude caused by the thermal plume to the self-stabilization capability of the adjacent units is calculated to obtain the influence degree coefficient matrix. The influence degree coefficient matrix characterizes the interference intensity of the thermal plume on the temperature field stability of each adjacent unit.

[0118] By combining the heat transfer capacity index and the influence coefficient matrix, the interference weight of each thermal plume convergence path in the overall temperature control process is calculated.

[0119] In this embodiment, a detailed geometric analysis and flow characteristic assessment are first performed on the identified thermal plume confluence paths. This analysis includes examining geometric features such as path length, curvature, cross-sectional changes, and bifurcation / merging. The complexity of the flow within the path is assessed, including the presence of phenomena such as separation, backflow, and secondary flow. A mesh generation method is used to set virtual detection planes at key locations along the thermal plume confluence paths. For example, the density of detection planes is increased in areas with drastic geometric changes and complex flows, while the density is appropriately reduced in relatively stable flow areas. This ensures that important flow characteristics are captured while controlling computational costs. The position and orientation of the detection planes are dynamically adjusted according to the flow characteristics of the thermal plume confluence paths, making the detection planes as perpendicular as possible to the mainstream direction to improve data acquisition accuracy. High-frequency time-series sampling (e.g., 10-100 times per second) is performed on each virtual detection plane, and the distribution of multiple physical quantities on the detection planes is recorded, including instantaneous data such as temperature field distribution (reflecting heat distribution), velocity vector field (reflecting momentum transfer), and pressure distribution (reflecting flow driving force). The changes in these instantaneous data over time are analyzed to identify periodic fluctuations, random fluctuations, and trends. The dynamic flow characteristic curves are constructed based on these changes, integrating spatiotemporal variation information of multiple physical quantities to comprehensively describe the dynamic transport characteristics of the heat plume. Then, based on the dynamic flow characteristic curves of each virtual detection plane, the instantaneous heat flow through the detection plane is calculated using the area integral method: Heat flow = ∫∫ρ×cp×v×TdA, where ρ is the fluid density, cp is the specific heat capacity, v is the flow velocity, and T is the absolute temperature. By integrating over time, the total heat transported along the confluence path of the heat plume within a specified time period is calculated, and the temporal variation law and periodic characteristics of heat transport are analyzed. Using fluid dynamics analysis methods, the heat distribution density of the path cross section is calculated, the uniformity and concentration of heat distribution on the cross section are analyzed, the velocity gradient distribution is calculated, and flow parameters such as boundary layer thickness and shear stress distribution are evaluated. The influence of velocity gradient on heat transfer is analyzed. Combining the geometric parameters (such as length, diameter, curvature, etc.) and thermal property parameters (such as fluid thermal conductivity, viscosity, etc.) of the heat plume confluence path, a quantitative model of heat transfer index is established. This index comprehensively reflects the heat transfer capacity and efficiency of the path, taking into account the combined effects of geometric effects, flow effects and thermal property effects.

[0120] Next, temperature monitoring was performed on the virtual thermal boundary of adjacent independent thermal control units to obtain high-resolution data on the boundary temperature changes over time. A multi-scale analysis method was used to decompose the temperature fluctuation signal. Wavelet transform was applied to decompose the temperature signal into components of different frequency scales, identifying short-term fluctuations (e.g., caused by equipment switching), medium-term fluctuations (e.g., caused by load changes), and long-term fluctuations (e.g., caused by environmental changes). Empirical Mode Decomposition (EMD) was used to extract the intrinsic mode functions (IMFs) of the temperature signal, analyzing the frequency and amplitude characteristics of each mode, identifying characteristic frequency components caused by the thermal plume, and establishing a mathematical model of the influence of the thermal plume on the temperature field of adjacent units. The model describes the transmission relationship between thermal plume disturbances and temperature response, considering factors such as the intensity, frequency, and duration of the disturbance, and analyzing the self-stabilization capability of each independent thermal control unit, including thermal inertia and control... The system's response capability and thermal buffering capability are assessed. The dynamic ratio of temperature disturbance intensity to unit self-stabilization capability is calculated. This ratio reflects the relative influence of thermal plumes on unit temperature control; a higher ratio indicates a more significant interference effect from the thermal plumes. Finally, based on the combined results of heat transfer index and temperature influence analysis, an evaluation index system for interference weights is established using the Analytic Hierarchy Process (AHP), including indicators for transfer capacity, influence intensity, duration, and influence range. A judgment matrix is ​​constructed to determine the relative importance weights of each indicator. Weight coefficients are determined through expert consultation and data analysis. Using a fuzzy comprehensive evaluation method, quantitative indicators and qualitative evaluations are combined to establish a fuzzy evaluation matrix, considering the uncertainty and fuzziness of the indicators. The comprehensive interference weight of each thermal plume convergence path is calculated, reflecting the importance of the path's interference throughout the temperature control process.

[0121] It should be further explained that, in the specific implementation process, the process of adjusting the convection path includes:

[0122] Heat overflow risk is identified based on a multi-source energy dynamic allocation matrix, resulting in independent thermal control units with heat overflow risk. Topological analysis is performed on the convection channels between multiple independent thermal control units to identify key convection nodes between each independent thermal control unit. Based on this, the convection path of the independent thermal control units with heat overflow risk is reconstructed, and the heat distribution is verified.

[0123] In this embodiment, the energy allocation ratio corresponding to each independent thermal control unit is obtained based on the multi-source energy dynamic allocation matrix, and the energy density corresponding to the corresponding independent thermal control unit, that is, the energy input or output per unit volume, is obtained based on it; and it is compared with a preset density threshold. If the corresponding energy density exceeds the preset density threshold, it is marked as a potential heat overflow area.

[0124] Furthermore, the temperature gradient deviation between the potential heat overflow area and adjacent independent thermal control units is assessed. If the temperature gradient deviation exceeds a preset deviation threshold and lasts for more than 10 seconds, it is marked as a heat overflow area, triggering a heat overflow warning. By identifying the direction and intensity of the heat overflow, heat overflow areas are classified into two types: cold overflow (diffusion from low temperature to high temperature) and hot overflow (diffusion from high temperature to low temperature). For cold overflow areas, the focus is primarily on energy waste; for hot overflow areas, the focus is on the impact on the quality of low-temperature products. Simultaneously, when a heat overflow warning is triggered, a virtual thermal boundary is defined to indicate the risk of heat overflow.

[0125] Based on the internal physical structure and ventilation layout of the vending machine, a convection channel topology network model is constructed. Each network node represents a thermal control unit or important airflow components (such as fans, evaporators, air outlets, etc.). Connecting edges represent possible airflow channels, and physical characteristic parameters, including cross-sectional area, length, curvature, and surface roughness, are assigned to each channel. The hydrodynamic drag coefficient of each channel is calculated to predict the airflow distribution under different fan operating conditions. Computational fluid dynamics simulation is used to verify the accuracy of the topology model and ensure that the model can accurately reflect the actual airflow organization.

[0126] Then, based on the constructed convection channel topology network model, the heat transfer paths between each independent thermal control unit are identified, and the contribution of each heat transfer path to heat overflow is evaluated based on the heat exchange capacity on the heat transfer path, and key convection nodes are selected based on it.

[0127] Furthermore, the topology network model of the convection channel is visualized to form a heat flow diagram, which intuitively shows the flow pattern of heat in the system. This topology-based analysis method can understand the heat transfer mechanism inside the vending machine from a holistic perspective, providing systematic guidance for subsequent convection path optimization.

[0128] Furthermore, based on the obtained heat flow diagram and key convection nodes, and combined with computational fluid dynamics theory, a three-dimensional dynamic model is constructed that includes the internal spatial geometry, boundary conditions, and heat source distribution of the vending machine. The three-dimensional dynamic model adopts the standard turbulence model.

[0129] A multi-objective optimization function is set up, which includes three aspects: temperature uniformity, energy efficiency, and convection intensity. Using this multi-objective optimization function as a constraint, a constructed three-dimensional dynamic model is used to simulate airflow and search for the optimal combination of convection path parameters. Based on the optimal combination of convection path parameters, the optimal angle of each air duct guide plate and the optimal speed of the fan in the vending machine's internal space are determined. An automated actuator is then constructed based on this, including a stepper motor-driven guide plate adjuster and a frequency converter-controlled fan driver. The air duct guide plates are precisely adjusted with an angle adjustment accuracy of ±2° to ensure airflow follows the designed path. The fan speed adjustment range is 30% to 100% of the rated speed, with a step size of 5%, enabling fine adjustment of airflow intensity.

[0130] In addition, differentiated adjustment strategies are adopted for different heat overflow situations; for cold overflow areas, the main approach is to reduce the convection intensity of the cold source area or redirect the cold airflow; for hot overflow areas, the main approach is to enhance the heat dissipation capacity of the heat source area or set up airflow barriers to isolate the heat flow.

[0131] It should be noted that the adjustment process is gradual, with each adjustment controlled within a small range to avoid drastic system fluctuations.

[0132] Once adjustments are complete, increase the sampling frequency of the deployed sensor array to capture rapid temperature changes; obtain the temperature change rate and gradient change of the heat overflow area before and after adjustments; and based on this, compare and analyze the energy balance state of each independent thermal control unit to verify the effectiveness of energy allocation; draw a three-dimensional thermal field distribution map after adjustments to visually display the improvement in heat distribution; calculate the heat overflow control effect evaluation indicators, including temperature deviation reduction rate, energy utilization efficiency improvement rate, and system stability index; when the temperature deviation reduction rate reaches more than 85% and the system stability index is less than 0.1, the heat overflow is considered to be effectively controlled; for areas with poor control effect, record the problem and trigger a secondary optimization process; generate a heat overflow control effect report, recording the adjustment process and effect evaluation results.

[0133] The response feedback module is used to continuously monitor the internal space of the vending machine after adjustment, and update the corresponding energy demand index in real time according to the frequency of micro-environmental temperature fluctuations in each independent thermal control unit.

[0134] It should be further explained that, during the specific implementation process, the internal space of the adjusted vending machine is continuously monitored to obtain the corresponding micro-environment temperature fluctuation frequency; and the process of updating the corresponding energy demand index in real time based on the micro-environment temperature fluctuation frequency includes:

[0135] Continuous multimodal data acquisition is performed on the internal space of the adjusted vending machine, and the temperature time series corresponding to each independent thermal control unit is obtained based on the data.

[0136] Frequency domain analysis was performed on the temperature time series of each independent thermal control unit to obtain the main frequency component and fluctuation amplitude of temperature fluctuations, and to obtain the frequency characteristics of microenvironment temperature fluctuations.

[0137] The frequency characteristics of temperature fluctuations are compared and analyzed with preset stability thresholds, and the temperature control effect and stability level of each independent thermal control unit are evaluated based on these thresholds to obtain a stability score. The corresponding energy demand index is then updated in real time based on the obtained stability score.

[0138] In this embodiment, the original sensor array is kept running continuously to obtain an adjusted temperature time series. This dataset reflects the actual environmental response after the control measures are implemented.

[0139] Then, the temperature time series collected in each independent thermal control unit is digitally processed, and the time domain signal is converted into the frequency domain signal through fast Fourier transform to obtain the spectral characteristics of temperature fluctuation, including the main frequency component, spectral energy distribution and bandwidth.

[0140] Based on the spectral characteristics, the periodicity and randomness of temperature fluctuations are obtained, the amplitude and root mean square value of temperature fluctuations are obtained, and based on these, the temperature fluctuation frequency feature vector F=[dominant frequency component, temperature fluctuation amplitude, root mean square value, bandwidth] is defined. This vector comprehensively describes the dynamic characteristics of the temperature environment within the unit.

[0141] Multi-level stability thresholds are set, including excellent stability (fluctuation amplitude <0.2℃, main frequency <0.01Hz), good stability (fluctuation amplitude <0.5℃, main frequency <0.02Hz), and acceptable stability (fluctuation amplitude <1.0℃, main frequency <0.05Hz). The real-time temperature fluctuation frequency feature vector is compared with the preset stability thresholds in multiple dimensions, and the temperature control effect of each independent thermal control unit is comprehensively evaluated using a preset neural network model to generate a stability score. Based on the stability score, units with poor temperature control effect and control parameters that need to be adjusted are identified.

[0142] Finally, parameters are dynamically adjusted based on the temperature stability assessment results. For example, for independent thermal control units with low stability scores, the weight of their energy demand index is increased, more cooling or heating resources are allocated preferentially, and the adjustment amount of the energy demand index is obtained according to the stability deviation. Where e represents the stability score bias, , , For control parameters;

[0143] Furthermore, the energy demand index of the corresponding independent thermal control unit is adjusted based on the adjustment amount, and the multi-source energy dynamic allocation matrix is ​​updated to re-optimize the coordinated operation ratio of each refrigeration device.

[0144] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

[0145] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0146] In the description of this invention, it should be understood that the terms "first," "second," etc., are used only for distinguishing descriptions and should not be construed as indicating or implying relative importance.

[0147] In the description of this invention, unless otherwise stated, "a plurality of" means two or more.

[0148] In the description of this invention, "several" means one or more, and "a large number" means two or more.

[0149] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0150] All formulas in this manual are dimensionless and calculated numerically. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters and thresholds in the formulas are set by those skilled in the art according to the actual situation.

[0151] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

Claims

1. An intelligent vending machine system with adaptive multi-zone temperature control function, characterized in that, include: The data acquisition module is used to collect data from the inside of the vending machine and the surface of the goods based on a preset sensor array, and to generate multimodal heat flux data. The map construction module is used to iteratively fit the real-time acquired multimodal heat flux data to obtain a three-dimensional thermal field distribution map. The region calibration module is used to perform real-time analysis based on the isothermal linear density and heat flow vector direction within the three-dimensional thermal field distribution map. Based on the real-time analysis results, the internal space of the vending machine is divided into several independent thermal control units. Virtual thermal boundaries are defined for adjacent independent thermal control units, and the confluence paths of heat plumes between adjacent independent thermal control units are identified based on the virtual thermal boundaries. The response generation module is used to obtain the energy demand index between the target temperature of each independent thermal control unit and the instantaneous three-dimensional thermal field distribution map, and combine the interference weight of the thermal plume confluence path to generate a multi-source energy dynamic allocation matrix. The response control module is used to identify the heat overflow risk of independent thermal control units based on the multi-source energy distribution matrix, and to adjust the convection path of independent thermal control units with heat overflow risk. The response feedback module is used to continuously monitor the internal space of the vending machine after adjustment, obtain the corresponding micro-environment temperature fluctuation frequency, and update the corresponding energy demand index in real time based on the micro-environment temperature fluctuation frequency.

2. The intelligent vending machine system with adaptive multi-zone temperature control function according to claim 1, characterized in that, The process of generating a three-dimensional thermal field distribution map includes: The collected multimodal heat flux data is spatially continuous to obtain the continuous temperature field distribution corresponding to the internal space of the vending machine, and the temperature gradient vector of each location in the internal space of the vending machine is obtained based on it. The heat transfer coefficients between products in the microenvironment are obtained based on a pre-built reference database of the thermal properties of the products, and the radiative heat flux data between the corresponding products in the microenvironment are obtained based on the heat transfer coefficients and temperature gradient vectors. The obtained radiative heat flux data is spatially fused with the continuous temperature field distribution, and the thermal field is reconstructed to obtain a three-dimensional thermal field distribution map.

3. The intelligent vending machine system with adaptive multi-zone temperature control function according to claim 2, characterized in that, The process of spatially continuum processing of the collected multimodal heat flux data includes: The collection location of contact temperature data in the vending machine's internal space is obtained from the corresponding multimodal heat flux data, and spatial gridding is performed on it to obtain the inner wall temperature distribution feature vector, which includes the temperature gradient change rate and the boundary temperature conduction coefficient. The temperature feature vector of the commodity surface is obtained based on infrared radiation data in multimodal heat flux data. The temperature feature vector includes surface temperature uniformity and hot spot distribution index. Data fusion and spatial interpolation are performed based on the obtained inner wall temperature distribution feature vector and the product surface temperature feature vector to obtain the continuous temperature field distribution corresponding to the vending machine content space.

4. The intelligent vending machine system with adaptive multi-zone temperature control function according to claim 2, characterized in that, The process of obtaining the heat transfer coefficient includes: Based on the thermal property parameters and surface area of ​​the commodity materials stored in the pre-built commodity thermal property reference database, the convective heat transfer coefficient between commodities is obtained; and the radiative heat transfer coefficient is obtained simultaneously; and a comprehensive heat transfer coefficient is obtained based on the convective heat transfer coefficient and the radiative heat transfer coefficient.

5. The intelligent vending machine system with adaptive multi-zone temperature control function according to claim 2, characterized in that, The process of obtaining the convergence path of thermal plumes includes: Isothermal surfaces are extracted from the three-dimensional thermal field distribution map to obtain isothermal surfaces at different temperature levels; and isothermal linear density indices are obtained based on the curvature characteristics and density distribution of the isothermal surfaces. Based on the temperature gradient vector in the three-dimensional thermal field distribution map, the divergence and curl of the heat flux vector direction are obtained, and the characteristic parameters of the heat flux vector are obtained. The coupling analysis of isothermal linear density index and heat flux vector characteristic parameters identifies spatial regions with similar heat transfer behavior, and based on this, the internal space of the vending machine is divided into several independent thermal control units. A virtual thermal boundary network is formed by connecting the temperature gradient abrupt change points at the boundaries of adjacent independent thermal control units; and by analyzing the temperature gradient discontinuity in the virtual thermal boundary network, the thermal plume convergence path between independent thermal control units is obtained.

6. The intelligent vending machine system with adaptive multi-zone temperature control function according to claim 5, characterized in that, The process of obtaining the multi-source energy dynamic allocation matrix includes: Set the target temperature range for each independent heat control unit, and determine the temperature control accuracy based on the preservation requirements of the goods in the area corresponding to each independent heat control unit; Based on the instantaneous three-dimensional thermal field distribution map, the deviation vector between the actual temperature distribution and the target temperature in each independent thermal control unit is obtained, and the energy change required for each independent thermal control unit to reach the target temperature is obtained based on it, thus obtaining the energy demand index. Flow analysis is performed on the thermal plume confluence path to obtain the heat transfer capacity of each path and the degree of impact on adjacent units, and the interference weight corresponding to the thermal plume confluence path is obtained based on it. Based on the energy demand index and corresponding disturbance weight of each independent thermal control unit, a multi-objective optimization problem is constructed and solved to obtain a multi-source energy dynamic allocation matrix.

7. The intelligent vending machine system with adaptive multi-zone temperature control function according to claim 6, characterized in that, The process of obtaining the interference weights includes: Multiple virtual detection planes are set up on the identified thermal plume confluence path, and time-series sampling is performed on the corresponding virtual detection planes to record the temperature distribution and velocity vector of each virtual detection plane at different time points. Based on this, a dynamic flow characteristic curve is constructed, which includes the heat flux change rate and directional stability index. The total heat transfer of the thermal plume confluence path per unit time is obtained based on the constructed dynamic flow characteristic curve, and the heat distribution density and velocity gradient of the path cross section are analyzed; combined with the geometry and transmission direction of the thermal plume confluence path, the heat transfer index of each thermal plume confluence path is obtained. Temperature fluctuation spectrum analysis is performed on the virtual thermal boundary corresponding to adjacent independent thermal control units to identify the temperature disturbance mode of the thermal plume on the adjacent independent thermal control units. Based on this, the ratio of the temperature fluctuation amplitude caused by the thermal plume to the self-stabilization capability of the adjacent units is calculated to obtain the influence degree coefficient matrix. The influence degree coefficient matrix characterizes the interference intensity of the thermal plume on the temperature field stability of each adjacent unit. By combining the heat transfer capacity index and the influence coefficient matrix, the interference weight of each thermal plume convergence path in the overall temperature control process is calculated.

8. The intelligent vending machine system with adaptive multi-zone temperature control function according to claim 7, characterized in that, The process of adjusting the convection path includes: Heat overflow risk is identified based on a multi-source energy dynamic allocation matrix, resulting in independent thermal control units with heat overflow risk. Topological analysis is performed on the convection channels between multiple independent thermal control units to identify key convection nodes between each independent thermal control unit. Based on this, the convection path of the independent thermal control units with heat overflow risk is reconstructed, and the heat distribution is verified.

9. The intelligent vending machine system with adaptive multi-zone temperature control function according to claim 8, characterized in that, The internal space of the adjusted vending machine was continuously monitored to obtain the corresponding micro-environment temperature fluctuation frequency. The process of updating the corresponding energy demand index in real time based on the frequency of microenvironment temperature fluctuations includes: Continuous multimodal data acquisition is performed on the internal space of the adjusted vending machine, and the temperature time series corresponding to each independent thermal control unit is obtained based on the data. Frequency domain analysis was performed on the temperature time series of each independent thermal control unit to obtain the main frequency component and fluctuation amplitude of temperature fluctuations, and to obtain the frequency characteristics of microenvironment temperature fluctuations. The frequency characteristics of temperature fluctuations are compared and analyzed with preset stability thresholds, and the temperature control effect and stability level of each independent thermal control unit are evaluated based on these thresholds to obtain a stability score. The corresponding energy demand index is then updated in real time based on the obtained stability score.

10. The intelligent vending machine system with adaptive multi-zone temperature control function according to claim 8, characterized in that, The process of identifying heat spillover risk based on a multi-source energy dynamic allocation matrix includes: The energy density of each independent thermal control unit is obtained and compared with a preset density threshold. Areas exceeding the threshold are marked as potential heat overflow areas. The temperature gradient deviation between the potential heat overflow area and the adjacent independent thermal control unit is evaluated. When the temperature gradient deviation exceeds a preset deviation threshold and the duration exceeds a preset duration, it is finally marked as an independent thermal control unit with a risk of heat overflow.

Images