A dew point dynamic tracking and double mode cooperative control based glass defogging intelligent system

Abstract

This invention relates to the field of industrial control system technology, specifically to an intelligent glass defogging system based on dew point dynamic tracking and dual-mode collaborative control. The system includes: an environmental parameter discrete module that collects temperature and relative humidity parameters from physical sensors and reads pulse-width modulation (PWM) heat integral parameters from the defogging heating hardware port; and constructs a two-dimensional thermal grid coordinate system using the physical sensor temperature parameters, the PWM heat integral parameters, preset glass heat capacity parameters, and area size parameters, calculating and generating a set of grid node temperatures. In this invention, large duty cycle control parameters or smooth duty cycle transition values ​​are output according to different operating states, enabling rapid improvement of heating capacity in the initial stage of defogging, and maintaining continuous output variation after the risk decreases. This ensures timely elimination of fog on the glass surface while reducing thermal shock and energy consumption fluctuations caused by sudden output changes, thereby improving the overall response speed of the glass defogging process.

Description

Technical Field

[0001] This invention relates to the field of industrial control system technology, and in particular to a smart glass defogging system based on dew point dynamic tracking and dual-mode collaborative control. Background Technology

[0002] Industrial control system technology is a cross-disciplinary technology that includes intelligent control of industrial processes, real-time monitoring of environmental parameters, embedded control, thermal management control, and intelligent decision control. It is mainly used for real-time perception, status analysis, control decision-making, and execution control of industrial equipment, industrial production objects, and industrial environment.

[0003] Existing industrial control systems for glass defogging mostly rely on real-time ambient temperature or relative humidity values ​​for control decisions. This reliance primarily reflects the environmental conditions at the sensor's location. However, heat transfer within the glass varies spatially, and the actual temperature changes in glass edges, localized heat dissipation areas, and areas with insufficient heating coverage are difficult to reflect synchronously. This can lead to situations where sensor readings meet control conditions, but certain areas remain close to the dew point, resulting in residual fog or even re-fogging on the glass surface. Therefore, improvements are needed. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of existing technologies and propose a smart glass defogging system based on dynamic dew point tracking and dual-mode collaborative control.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a smart glass defogging system based on dew point dynamic tracking and dual-mode collaborative control includes: The environmental parameter discrete module collects temperature and relative humidity parameters from physical sensors and reads pulse width modulation heat integration parameters from the defogging heating hardware port. It constructs a two-dimensional thermal grid coordinate system using the physical sensor temperature parameters, the pulse width modulation heat integration parameters, and preset glass heat capacity and area size parameters, and calculates and generates a set of grid node temperatures. It extracts the calculated temperature corresponding to the lowest value node within the grid node temperature set and combines it with the relative humidity parameter to generate extreme value node state coordinates. The enthalpy-humidity matrix addressing module inputs the extreme node state coordinates into a multidimensional digital enthalpy-humidity matrix to generate a thermodynamic state mapping value; it calculates the spatial vector difference between the thermodynamic state mapping value and the dew point critical curve parameters pre-stored in the control program to generate a critical safety vector distance. The fuzzy inference debouncing module inputs the critical safety vector distance into the fuzzy logic program layer to generate fuzzy membership degree values; generates short-distance trigger limits and long-distance exit limits based on the fuzzy membership degree values, and merges them to construct a dual switching boundary; and generates a dual-mode state switching command based on the comparison result between the critical safety vector distance and the dual switching boundary values. The dual-mode collaborative output module reads the dual-mode state switching instruction. When a high-efficiency defogging indicator is obtained, it generates a large duty cycle control parameter. When a high-efficiency maintenance indicator is obtained, it calculates and generates a duty cycle smooth transition value based on the current duty cycle parameter. The large duty cycle control parameter or the duty cycle smooth transition value is converted into pulse width modulation waveform data to drive the glass defogging hardware to defog.

[0006] Preferably, the step of obtaining the temperature set of the grid nodes is as follows: Within the same sampling period, the temperature and relative humidity parameters of the physical sensor are collected. The time correspondence between the temperature and relative humidity parameters of each group of physical sensor is recorded according to the sampling time. The pulse width modulation heat integration parameters of the defogging heating hardware port within the same sampling period are read. The preset glass heat capacity parameters and area size parameters are called. The grid row boundary, grid column boundary, number of nodes and node spatial position are determined according to the area size parameters. The heat capacity attribute is configured for each node according to the glass heat capacity parameters. The physical sensor temperature parameters are written to the glass boundary nodes, and the pulse width modulation heat integration parameters are written to the boundary nodes corresponding to the heating port to form a two-dimensional thermal grid coordinate. Based on the coordinates of the two-dimensional thermal grid, the row and column positions, heat capacity attributes, boundary temperatures, and heating boundary heat of each node are read row by row. According to the spatial position of the nodes, the adjacent nodes of each internal node in the row front, row back, column front, and column back are determined. The difference between the estimated temperature of each adjacent node and the estimated temperature of the current node is converted into row-direction heat change and column-direction heat change according to the node spatial interval. The estimated temperature of the current node is updated by combining the heat capacity attribute of the current node and the heating boundary heat of the current iteration cycle. After completing one round of updates, the estimated temperature difference of nodes in two adjacent rounds is compared one by one. When the estimated temperature difference of all nodes does not exceed the preset convergence limit, the update stops. When the estimated temperature difference of any node exceeds the preset convergence limit, the next round of updates continues. The estimated temperatures of each node when the update stops are summarized to generate a set of grid node temperatures.

[0007] Preferably, the step of obtaining the extreme value node state coordinates is as follows: Based on the set of grid node temperatures, the estimated temperatures of each node are read sequentially according to a fixed row and column order. The estimated temperature of the first node is registered as the current minimum temperature. The estimated temperatures of subsequent nodes are compared with the current minimum temperature. When the estimated temperature of a subsequent node is less than the current minimum temperature, the current minimum temperature is updated and the corresponding node is registered. When the estimated temperature of a subsequent node is not less than the current minimum temperature, the current minimum temperature and the corresponding node remain unchanged. After traversing all nodes, the node with the lowest value is determined. The estimated temperature corresponding to the node with the lowest value and the relative humidity parameter at the same sampling time are extracted. The estimated temperature is converted into a matrix row index value according to a preset temperature index interval, and the relative humidity parameter is converted into a matrix column index value according to a preset humidity index interval. When the estimated temperature or relative humidity parameter is located at the boundary of an adjacent index interval, the absolute difference between the parameter value and the center value of the adjacent interval is compared. The interval number with the smallest absolute difference is selected as the corresponding index value. The matrix row index value and the matrix column index value are combined to generate the state coordinates of the extreme value node.

[0008] Preferably, the step of obtaining the thermodynamic state mapping value is as follows: The system invokes a pre-stored multidimensional digital enthalpy-humidity matrix in the microprocessor memory, reads the row index range, column index range, dimensional arrangement order, and intersection storage location corresponding to the multidimensional digital enthalpy-humidity matrix, sequentially extracts the matrix row index value and matrix column index value from the extreme value node state coordinates, matches the matrix row index value with the row index range of the multidimensional digital enthalpy-humidity matrix item by item, and matches the matrix column index value with the column index range of the multidimensional digital enthalpy-humidity matrix item by item. When both the matrix row index value and the matrix column index value fall within the corresponding index range, the row and column intersection point pointed to by the matrix row index value and the matrix column index value are locked, the thermodynamic data stored in the row and column intersection point according to the dimensional arrangement order is read, and the thermodynamic data is combined according to the original dimensional order to generate a thermodynamic state mapping value.

[0009] Preferably, the step of obtaining the critical safety vector distance is as follows: The system reads the pre-stored dew point critical curve parameters from the control program, parses the curve node sequence, state dimension identifier, and component values ​​of each curve node corresponding to the dew point critical curve parameters, expands the thermodynamic state mapping value into a state component sequence according to the state dimension identifier, arranges the components of each curve node in the dew point critical curve parameters according to the same state dimension identifier, compares the state component sequence of the thermodynamic state mapping value with the component arrangement order of each curve node, excludes curve nodes with inconsistent state dimension identifiers, calculates the component value difference of each dimension for curve nodes with consistent state dimension identifiers, records the component value difference and positive and negative directions of each component according to the curve node sequence, and generates a spatial vector difference value. Based on the spatial vector difference, the component value difference corresponding to each curve node is read in groups. The component value difference of each dimension under the same curve node is squared. All squared results under the same curve node are accumulated. The accumulated results are square rooted to obtain the vector distance value corresponding to each curve node. All vector distance values ​​are compared in the order of curve nodes of the dew point critical curve parameter. The first vector distance value is registered as the current minimum distance. When the subsequent vector distance value is less than the current minimum distance, the current minimum distance is updated. When the subsequent vector distance value is not less than the current minimum distance, the current minimum distance is kept unchanged. After completing the comparison of all vector distance values, the critical safe vector distance is generated.

[0010] Preferably, the step of obtaining the fuzzy membership degree value is as follows: The critical safety vector distance is input into the fuzzy logic program layer. The pre-stored distance segment boundaries, membership interval endpoints, interval increasing direction, and interval decreasing direction within the fuzzy logic program layer are read. The corresponding distance segments are retrieved according to the numerical position of the critical safety vector distance. The membership interval endpoints corresponding to both ends of the distance segments are extracted. The interval position ratio of the critical safety vector distance relative to the starting boundary of the distance segment is calculated. The numerical proportion of the membership interval endpoints is adjusted according to the interval position ratio. When the critical safety vector distance is located at the common boundary of adjacent distance segments, the membership values ​​corresponding to the adjacent distance segments are calculated respectively. The membership value with the largest value is selected to generate the fuzzy membership degree value.

[0011] Preferably, the step of obtaining the dual-mode state switching instruction is as follows: The system reads the pre-stored trigger limit benchmark, exit limit benchmark, trigger limit adjustment range, and exit limit adjustment range within the control program. Based on the numerical position of the fuzzy membership degree value in the membership interval, it determines the reduction range corresponding to the trigger limit benchmark and the expansion range corresponding to the exit limit benchmark. It subtracts the reduction range from the trigger limit benchmark to generate a short-distance trigger limit. It adds the expansion range to the exit limit benchmark to generate a long-distance exit limit. It compares the short-distance trigger limit and the long-distance exit limit. When the short-distance trigger limit is not less than the long-distance exit limit, it lowers the short-distance trigger limit according to a preset boundary interval and raises the long-distance exit limit according to the same preset boundary interval. It completes the combination in the order of short-distance trigger limit first and long-distance exit limit second, forming a dual switching boundary. Based on the aforementioned dual switching boundaries, short-distance trigger limits and long-distance exit limits are extracted. The critical safety vector distance is compared numerically with both the short-distance trigger limit and the long-distance exit limit. When the critical safety vector distance is less than the short-distance trigger limit, a trigger mode switching flag is written, and the continuous dwell count is cleared. When the critical safety vector distance is between the short-distance trigger limit and the long-distance exit limit, the current state machine position is maintained, and the continuous dwell count is cleared. When the critical safety vector distance is greater than the long-distance exit limit, the continuous dwell count is incremented sequentially according to machine cycles. When the continuous dwell count does not exceed the preset number of machine cycles, the current state machine position is maintained. When the continuous dwell count exceeds the preset number of machine cycles, an exit condition fulfillment flag is written. The corresponding state machine target position is selected according to the trigger mode switching flag or the exit condition fulfillment flag, and the state machine is controlled to jump to either the high-efficiency defogging state or the high-efficiency maintenance state, generating a dual-mode state switching command.

[0012] Preferably, the specific steps for driving the glass defogging hardware are as follows: Read the dual-mode state switching instruction, extract the state identifier field, switching effect field, instruction cycle field, and state jump target field according to the order of instruction fields, match the state identifier field bit by bit with the strong defogging state code and the high energy efficiency maintenance state code, check whether the switching effect field is in the allowed output state, check whether the state jump target field corresponds to the strong defogging state or the high energy efficiency maintenance state, when the state identifier field matches the strong defogging state code and the state jump target field corresponds to the strong defogging state, register the strong defogging identifier, when the state identifier field matches the high energy efficiency maintenance state code and the state jump target field corresponds to the high energy efficiency maintenance state, register the high energy efficiency maintenance identifier, and obtain the strong defogging identifier or the high energy efficiency maintenance identifier; When registering the high-efficiency defogging flag, the current duty cycle parameter, the target duty cycle parameter for high-efficiency defogging, and the upper limit of the allowable duty cycle for the defogging heating hardware port are read. The target duty cycle parameter for high-efficiency defogging is compared with the upper limit of the allowable duty cycle. The target duty cycle parameter for high-efficiency defogging that does not exceed the upper limit of the allowable duty cycle is selected and written into the high-efficiency defogging output field to generate a large duty cycle control parameter. When registering the high-efficiency maintenance flag, the current duty cycle parameter, the target duty cycle parameter for high-efficiency maintenance, the transition cycle length, and the control point arrangement parameter are read. The current duty cycle parameter is used as the transition start point, and the target duty cycle parameter for high-efficiency maintenance is used as the transition end point. The continuous output positions are divided according to the transition cycle length. The intermediate control points are adjusted according to the change direction of adjacent control points. When the change slope of adjacent output positions is discontinuous, the corresponding intermediate control points are readjusted until the change slope of adjacent output positions is continuous, generating a smooth transition value for the duty cycle. The pulse width modulation carrier period, timing count upper limit, output effective level, and defogging heating hardware port channel address are read. The large duty cycle control parameters are converted into effective level duration count values ​​within a single carrier period. The duty cycle smooth transition value is sequentially converted into effective level duration count values ​​for the corresponding carrier period according to the output position within the transition period. The ineffective level duration count value for each carrier period is determined according to the timing count upper limit. The effective level duration segment and ineffective level duration segment are arranged according to the output effective level. The carrier period, effective level duration count value, ineffective level duration count value, and defogging heating hardware port channel address are written into the pulse width modulation output register location to drive the glass defogging hardware to perform defogging.

[0013] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In this invention, temperature parameters from physical sensors, relative humidity parameters, and pulse width modulation heat integration parameters are used together as the basis for analyzing the thermal state of the glass. A two-dimensional thermal grid coordinate system is established by combining glass heat capacity parameters and area size parameters. This allows for continuous calculation of temperature changes in different areas of the glass. Then, the lowest value nodes are extracted from the grid node temperature set to construct extreme value node state coordinates. This transforms the control basis from a single-point environmental state to the actual thermal state of the location where the glass is most prone to fogging, thus more accurately reflecting the fogging risk on the glass surface. Furthermore, the extreme value node state coordinates are used to complete thermodynamic state mapping, and the critical safety vector distance is calculated by combining it with dew point critical curve parameters. This establishes a continuous quantitative relationship between the environmental state and the dew point boundary, avoiding the limitations of traditional control methods. Slight fluctuations in environmental parameters can lead to unstable risk assessment. Subsequently, fuzzy membership values ​​are generated based on the critical safety vector distance, and short-distance trigger limits and long-distance exit limits are dynamically formed. During mode switching, continuous state determination is added, ensuring that the control state only switches when the risk continuously changes, reducing control jitter caused by repeated entry and exit from defogging control near the critical state. After completing mode determination, large duty cycle control parameters or duty cycle smooth transition values ​​are output according to different operating states. This allows for rapid improvement of heating capacity in the initial stage of defogging, and maintains continuous output variation after the risk decreases. This ensures timely removal of fog from the glass surface while reducing thermal shock and energy consumption fluctuations caused by sudden output changes, thereby improving the overall response speed of the glass defogging process. Attached Figure Description

[0014] Figure 1 A schematic diagram of the temperature set of grid nodes; Figure 2 This is a schematic diagram of the critical safety vector distance and the dual switching boundary. Detailed Implementation

[0015] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0016] Please see Figure 1-2 This invention provides a technical solution: a smart glass defogging system based on dew point dynamic tracking and dual-mode collaborative control, comprising: The environmental parameter discrete module collects temperature and relative humidity parameters from physical sensors and reads pulse width modulation heat integration parameters from the defogging heating hardware port. It constructs a two-dimensional thermal grid coordinate system using the physical sensor temperature parameters, pulse width modulation heat integration parameters, preset glass heat capacity parameters, and area size parameters, and calculates and generates a set of grid node temperatures. It extracts the estimated temperature corresponding to the lowest value node in the grid node temperature set and converts it together with the relative humidity parameter to generate extreme value node state coordinates. The enthalpy-humidity matrix addressing module inputs the extreme node state coordinates into the multidimensional digital enthalpy-humidity matrix to generate thermodynamic state mapping values; it calculates the spatial vector difference between the thermodynamic state mapping values ​​and the dew point critical curve parameters pre-stored in the control program to generate the critical safety vector distance. The fuzzy inference debouncing module inputs the critical safety vector distance into the fuzzy logic program layer to generate fuzzy membership degree values; it generates short-distance trigger limits and long-distance exit limits based on the fuzzy membership degree values, and merges them to construct a dual switching boundary; it generates a dual-mode state switching command based on the comparison result of the critical safety vector distance and the dual switching boundary values. The dual-mode collaborative output module reads the dual-mode state switching command. When it obtains the high-efficiency defogging indicator, it generates a large duty cycle control parameter. When it obtains the high-efficiency maintenance indicator, it calculates and generates a duty cycle smooth transition value based on the current duty cycle parameter. It converts the large duty cycle control parameter or the duty cycle smooth transition value into pulse width modulation waveform data to drive the glass defogging hardware to defog.

[0017] The steps to obtain the temperature set of the grid nodes are as follows: Within the same sampling period, the temperature and relative humidity parameters of the physical sensor are collected. The time correspondence between the temperature and relative humidity parameters of each group of physical sensor is recorded according to the sampling time. The pulse width modulation heat integration parameters of the defogging heating hardware port within the same sampling period are read. The preset glass heat capacity parameters and area size parameters are called. The grid row boundary, grid column boundary, number of nodes and node spatial position are determined according to the area size parameters. The heat capacity attribute is configured for each node according to the glass heat capacity parameters. The physical sensor temperature parameters are written to the glass boundary nodes, and the pulse width modulation heat integration parameters are written to the boundary nodes corresponding to the heating port to form a two-dimensional thermal grid coordinate. Based on the coordinates of a two-dimensional thermal grid, the row and column positions, heat capacity attributes, boundary temperatures, and heating boundary heat of each node are read row by row. According to the spatial position of the nodes, the adjacent nodes of each internal node in the row front, row back, column front, and column back are determined. The difference between the estimated temperature of each adjacent node and the estimated temperature of the current node is converted into row-direction heat change and column-direction heat change according to the node spatial interval. The estimated temperature of the current node is updated by combining the heat capacity attribute of the current node and the heating boundary heat of the current iteration cycle. After completing one round of updates, the estimated temperature difference between adjacent rounds of nodes is compared one by one. When the estimated temperature difference of all nodes does not exceed the preset convergence limit, the update stops. When the estimated temperature difference of any node exceeds the preset convergence limit, the next round of updates continues. The estimated temperatures of each node at the time when the update stops are summarized to generate a set of grid node temperatures.

[0018] Specifically, within the same sampling period, the temperature and relative humidity parameters of the physical sensors are collected. The time correspondence between the temperature and relative humidity parameters of each set of physical sensors is recorded according to the sampling time. The pulse width modulation heat integral parameters of the defogging heating hardware port within the same sampling period are read. Preset glass heat capacity and area parameters are then retrieved. For example, for a laminated glass with an area of ​​1.2 square meters (1.5 meters × 0.8 meters) and a thickness of 5 millimeters, its preset comprehensive heat capacity parameter is 2.1 kJ / (kg·K), and its density is 2500 kg / m³. 3 Based on the area size parameter, a 150×80 two-dimensional grid matrix is ​​constructed at the logical level, with the grid row boundary set to 150, the grid column boundary set to 80, and the number of nodes arranged to 12,000. The spatial coordinates of each node are then calculated. The mapping relationship with physical dimensions, where The range is from 1 to 150. The range is 1 to 80. The heat capacity properties of each node are calculated based on the glass heat capacity parameters and the micro-element volume represented by each grid node. The temperature collected by physical sensors, such as the temperature values ​​measured by four temperature sensors installed at the four corners of the glass, is calculated using a bilinear interpolation algorithm and assigned to the four corner points and the corresponding boundary nodes of the grid. For the pulse width modulation heat integration parameter, it is first located to the corresponding boundary node according to the heating wire layout (e.g., serpentine layout), and then converted into the actual input heat using a formula. The calculation method is as follows ,in It is the rated power of the heating hardware, such as 100 watts. yes The pulse width modulation duty cycle at any given moment. It is the sampling period duration, for example, 1 second, which will be calculated. The heat flow boundary conditions are written into the boundary nodes corresponding to the heating ports to form two-dimensional thermal grid coordinates.

[0019] Based on the two-dimensional thermal grid coordinates, the row and column positions of each node are read row by row. Heat capacity properties The assigned boundary node temperatures and heating boundary heat are determined through a discretized solution of the two-dimensional unsteady-state heat conduction partial differential equation, specifically using the finite difference method to determine the internal nodes according to their spatial locations. On the front side of the line , walk to the back , column forward side and column to the rear Using Fourier's law of heat conduction, the temperature of each adjacent node is calculated. The difference in temperature relative to the current node, calculated according to the node spatial interval. and The change in heat flowing into or out of the current node is converted into the heat capacity attribute of the current node and the heating boundary heat of the current iteration period. The current node is then updated for the next time step using the following iterative formula. Calculated temperature : ,in, It is a node In the The temperature value of the next iteration, It is the thermal diffusivity of glass, and its value is ,in The thermal conductivity of glass is approximately 1.0 W / (m·K). For density, For specific heat capacity, The iteration time step, Spatial step size (e.g.) After completing one round of updates for all internal nodes, the absolute value of the temperature difference is calculated by comparing the nodes from two adjacent rounds. Preset convergence limit The setting is based on a balance between control accuracy requirements and computational resources. For example, setting it to 0.01℃ is much smaller than the resolution of the temperature sensor itself. When the absolute value of the temperature difference at all nodes is less than... When the calculation is considered convergent, the iterative process stops, and the absolute value of the calculated temperature difference at any node is not less than [a certain value]. When the convergence condition is met, the next round of iteration updates continues until the convergence condition is met. Finally, the final node estimated temperature when each node stops updating is summarized to generate a set of grid node temperatures.

[0020] The steps to obtain the state coordinates of extreme nodes are as follows: Based on the set of grid node temperatures, the estimated temperature of each node is read in a fixed row and column order. The estimated temperature of the first node is recorded as the current minimum temperature. The estimated temperatures of subsequent nodes are compared with the current minimum temperature. When the estimated temperature of a subsequent node is less than the current minimum temperature, the current minimum temperature is updated and the corresponding node is recorded. When the estimated temperature of a subsequent node is not less than the current minimum temperature, the current minimum temperature and the corresponding node remain unchanged. After traversing all nodes, the node with the lowest value is determined. The estimated temperature corresponding to the node with the lowest value and the relative humidity parameter at the same sampling time are extracted. The estimated temperature is converted into a matrix row index value according to the preset temperature index interval, and the relative humidity parameter is converted into a matrix column index value according to the preset humidity index interval. When the estimated temperature or relative humidity parameter is located at the boundary of an adjacent index interval, the absolute difference between the parameter value and the center value of the adjacent interval is compared. The interval number with the smallest absolute difference is selected as the corresponding index value. The matrix row index value and the matrix column index value are combined to generate the state coordinates of the extreme value node.

[0021] Specifically, based on the grid node temperature set, the calculated temperature of each node is read sequentially from row 1 to row 150, and within each row, from column 1 to column 80, in a fixed order. The first node... The estimated temperature is recorded as the current minimum temperature, and its node coordinates are recorded. Then, the estimated temperatures of subsequent nodes are read sequentially and compared with the current minimum temperature. When a subsequent node, such as node... If the calculated temperature is lower than the current minimum temperature, then the current minimum temperature is updated to the calculated temperature of that node, and the recorded node coordinates are updated synchronously. When the calculated temperature of a subsequent node is not less than the current minimum temperature, the current minimum temperature and the corresponding node coordinates remain unchanged until all 12,000 nodes have been traversed. The final recorded temperature and coordinates are the lowest value node and its calculated temperature. The calculated temperature corresponding to the lowest value node is extracted, for example, 12.3℃, and the ambient relative humidity parameter collected by the physical sensor at the same sampling time, for example, 78%. Then, these two physical quantities are converted into matrix row and column indices according to the preset temperature index interval and humidity index interval. The preset temperature index interval is set to [-10℃, 50℃], with a step size of 0.5℃, for a total of 120 intervals. The preset humidity index interval is [0%, 100%], with a step size of 1%, for a total of 100 intervals. For 12.3℃, it falls within the interval [12.0, 12.5), corresponding to the 45th temperature index, so the matrix row index value is 45. For 78%, it falls within the interval [78, Within 79), corresponding to the 79th humidity index, the matrix column index value is 79. When the calculated temperature or relative humidity parameter is exactly at the boundary of the adjacent index interval, for example, the temperature is 12.5℃, the absolute difference from the center value of the two adjacent intervals (12.25℃ and 12.75℃) is calculated respectively. The interval number with the smallest absolute difference is selected as the index. The final determined matrix row index value and matrix column index value are combined to generate the extreme value node state coordinates.

[0022] The steps for obtaining the thermodynamic state mapping value are as follows: The system calls upon a pre-stored multidimensional digital enthalpy-humidity matrix in the microprocessor's memory, reads the corresponding row index range, column index range, dimensional arrangement order, and intersection storage location of the multidimensional digital enthalpy-humidity matrix, sequentially extracts the matrix row index value and matrix column index value from the extreme node state coordinates, matches the matrix row index value with the row index range of the multidimensional digital enthalpy-humidity matrix item by item, and matches the matrix column index value with the column index range of the multidimensional digital enthalpy-humidity matrix item by item. When both the matrix row index value and matrix column index value fall within the corresponding index range, the system locks the row and column intersection point pointed to by the matrix row index value and matrix column index value, reads the thermodynamic data stored in the row and column intersection point according to the dimensional arrangement order, and combines the thermodynamic data according to the original dimensional order to generate thermodynamic state mapping values.

[0023] Specifically, the pre-stored multidimensional digital enthalpy-humidity matrix in the microprocessor memory is invoked. This matrix is ​​a discretized representation of the thermodynamic properties of air under standard atmospheric pressure. It is pre-calculated using the theoretical formula for air-water vapor mixtures provided by industry standards. The row index corresponds to the dry-bulb temperature, and the column index corresponds to the relative humidity. The stored dimensional data includes specific enthalpy, moisture content, specific volume, dew point temperature, etc. The range of row indexes (e.g., 1 to 121 corresponding to [-10℃, 50℃]), the range of column indexes (e.g., 1 to 101 corresponding to [0%, 100%]), the dimensional arrangement order (e.g., the first dimension stores specific enthalpy, the second dimension stores moisture content), and the memory base address of the intersection point storage location corresponding to the multidimensional digital enthalpy-humidity matrix are read. The matrix row index value (e.g., 45) and the matrix column index value (e.g., 79) are extracted sequentially from the extreme value node state coordinates generated in the previous step. The matrix row index value 45 is then compared with the row index range of the enthalpy-humidity matrix [1,

[121] Matching is performed by matching the matrix column index value 79 with the column index range [1, 101] of the enthalpy-humidity matrix. When both the matrix row index value and the matrix column index value fall within their respective valid index ranges, the absolute address of the target data in memory is calculated based on the memory base address, row index, column index, and the data byte length of each intersection point. The row and column intersection point is locked. For example, address = base address + (row index - 1) × (number of columns × number of data point bytes) + (column index - 1) × number of data point bytes. Starting from this address, the stored thermodynamic data is read continuously according to the pre-stored dimensional arrangement order. For example, 4 bytes of specific enthalpy floating-point number and 4 bytes of moisture content floating-point number are read. The read thermodynamic data is combined according to the original dimensional order to form a vector containing multiple thermodynamic parameters and generate thermodynamic state mapping values.

[0024] The steps to obtain the critical safety vector distance are as follows: The system reads the pre-stored dew point critical curve parameters from the control program, parses the curve node sequence, state dimension identifier, and component values ​​of each curve node corresponding to the dew point critical curve parameters, expands the thermodynamic state mapping value into a state component sequence according to the state dimension identifier, arranges the components of each curve node in the dew point critical curve parameters according to the same state dimension identifier, compares the state component sequence of the thermodynamic state mapping value with the component arrangement order of each curve node, eliminates curve nodes with inconsistent state dimension identifiers, calculates the component value difference of each dimension for curve nodes with consistent state dimension identifiers, records the component value difference and positive and negative directions of each component according to the curve node sequence, and generates a spatial vector difference value. Based on spatial vector differences, the component numerical differences corresponding to each curve node are read in groups. The numerical differences of each dimension of the components under the same curve node are squared respectively. All squared results under the same curve node are accumulated. The accumulated results are square rooted to obtain the vector distance values ​​corresponding to each curve node. All vector distance values ​​are compared according to the curve node order of the dew point critical curve parameters. The first vector distance value is registered as the current minimum distance. When the subsequent vector distance values ​​are less than the current minimum distance, the current minimum distance is updated. When the subsequent vector distance values ​​are not less than the current minimum distance, the current minimum distance is kept unchanged. After completing the comparison of all vector distance values, the critical safe vector distance is generated.

[0025] Specifically, the pre-stored dew point critical curve parameters are read from the control program. These parameters represent a saturation curve on a standard enthalpy-humidity chart, signifying 100% relative humidity. They consist of a series of discrete curve nodes, calculated and pre-stored according to the ASHRAE thermodynamic formula, used to define condensation and non-condensation states. The dew point critical curve parameters are analyzed to obtain the order of the curve nodes, the state dimension identifier of each node (e.g., dimension 1 is dry-bulb temperature, dimension 2 is moisture content), and the specific component values ​​of each node in these dimensions. The thermodynamic state mapping values ​​generated in the previous step are expanded into a state component sequence according to the same state dimension identifier. For example, if the mapping value is {specific enthalpy: h_current, moisture content: w_current, dry-bulb temperature: T_current}, and the curve node dimension identifier is {dry-bulb temperature, moisture content}, then the state component sequence is: Subsequently, each curve node in the dew point critical curve parameters is arranged according to the same state dimension identifier, forming a node sequence. Each node The process involves comparing the state component sequence of the thermodynamic state mapping values ​​with the component arrangement order of each curve node, eliminating curve nodes with inconsistent state dimension identifiers due to data definition errors or version mismatches, ensuring that the comparison is performed within the same physical dimension space. For all curve nodes with consistent state dimension identifiers, the component value difference between the current state point and that curve node is calculated dimension by dimension. and Record the numerical difference of each component according to the inherent order of the curve nodes. It also includes the positive and negative directions, generating spatial vector differences.

[0026] Based on spatial vector differences, the set of component numerical differences corresponding to each dew point critical curve node is read group by group. The components are dimensionless by dividing the difference between them by the typical range of their corresponding physical quantities. For example, the temperature range is taken as 50℃ and the moisture content range as 20g / kg, resulting in the normalized component differences. and Then, for the same curve node The differences in the normalized component values ​​of each dimension are squared, summed, and then the square root of the sum is applied. This is equivalent to calculating the normalized Euclidean distance from the current state point to each curve node, and the calculation formula is as follows: ,in, It is the current state point to the th The vector distance values ​​between the curve nodes. and These are the normalized temperature component difference and humidity component difference, respectively. This method is used to calculate the distance from the current state point to all... The set of vector distance values ​​for each critical curve node Compare all vector distance values ​​according to the curve node order of the dew point critical curve parameters, and select the first vector distance value. Register as the current minimum distance, then... Compared with the current minimum distance, if If it is smaller, update the current minimum distance to Otherwise, it remains unchanged. By traversing all vector distance values, the final minimum distance is the closest distance between the current air state point and the dew point critical curve, thus generating the critical safe vector distance.

[0027] The steps to obtain fuzzy membership values ​​are as follows: The critical safety vector distance is input into the fuzzy logic program layer. The pre-stored distance segment boundaries, membership interval endpoints, interval increasing direction, and interval decreasing direction are read from the fuzzy logic program layer. The corresponding distance segment is retrieved according to the numerical position of the critical safety vector distance. The membership interval endpoints corresponding to the two ends of the distance segment are extracted. The interval position ratio of the critical safety vector distance relative to the starting boundary of the distance segment is calculated. The numerical proportion of the membership interval endpoints is adjusted according to the interval position ratio. When the critical safety vector distance is located on the common boundary of adjacent distance segments, the membership values ​​corresponding to the adjacent distance segments are calculated respectively. The membership value with the largest value is selected to generate the fuzzy membership degree value.

[0028] Specifically, the critical safety vector distance is input into the fuzzy logic program layer. The layer then reads the pre-stored fuzzy set definition describing the condensation risk level, including distance segment boundaries, membership function types, and endpoints. For example, three fuzzy sets are preset: "Safe," "Critical," and "Dangerous." Their distance segment boundaries are set based on experience and extensive experimental data. For instance, "Safe" corresponds to a distance greater than 8, "Critical" to a distance between 3 and 10, and "Dangerous" to a distance less than 5. The membership function uses trapezoidal or triangular functions, and the increasing or decreasing direction of each interval is defined. For example, for an input critical safety vector distance value... Its membership degree in the "dangerous" fuzzy set The calculation process is as follows: when hour, ,when When the membership degree decreases linearly, its calculation formula is: ,when hour, The input distance is calculated based on this. Similarly, for the membership degree values ​​of the "dangerous" fuzzy set, calculate its membership degree values ​​for the "safe" and "critical" fuzzy sets. When the critical safety vector distance lies on the common boundary of adjacent distance segments, for example... It belongs to both the "critical" and "dangerous" fuzzy sets. In this case, it is necessary to calculate its membership degree in each set separately. Meanwhile, for example, the membership function of the "critical" fuzzy set in the interval [3, 10] is in the segment [3, 6]. ,but According to the rules of fuzzy inference, the fuzzy set with the largest membership degree is usually selected as the current state. However, in this step, it is only necessary to output the membership degree value that is most relevant to the risk, that is, select the membership value with the largest value and generate the fuzzy membership degree value.

[0029] The steps for obtaining the dual-mode state switching command are as follows: The system reads the pre-stored trigger limit reference, exit limit reference, trigger limit adjustment range, and exit limit adjustment range in the control program. Based on the numerical position of the fuzzy membership degree value in the membership interval, it determines the reduction range corresponding to the trigger limit reference and the expansion range corresponding to the exit limit reference. It subtracts the reduction range from the trigger limit reference to generate the short-distance trigger limit and adds the expansion range to the exit limit reference to generate the long-distance exit limit. It compares the short-distance trigger limit and the long-distance exit limit. When the short-distance trigger limit is not less than the long-distance exit limit, it lowers the short-distance trigger limit according to the preset boundary interval and raises the long-distance exit limit according to the same preset boundary interval. It completes the combination in the order of short-distance trigger limit first and long-distance exit limit second to form a dual switching boundary. Based on dual switching boundaries, short-distance trigger limits and long-distance exit limits are extracted. The critical safety vector distance is compared with the short-distance trigger limit and long-distance exit limit respectively. When the critical safety vector distance is less than the short-distance trigger limit, a trigger mode switching flag is written and the continuous dwell count is cleared to zero. When the critical safety vector distance is between the short-distance trigger limit and the long-distance exit limit, the current state machine position is maintained and the continuous dwell count is cleared to zero. When the critical safety vector distance is greater than the long-distance exit limit, the continuous dwell count is increased sequentially according to the machine cycle. When the continuous dwell count does not exceed the preset number of machine cycles, the current state machine position is maintained. When the continuous dwell count exceeds the preset number of machine cycles, an exit condition fulfillment flag is written. The corresponding state machine target position is selected according to the trigger mode switching flag or the exit condition fulfillment flag, and the state machine is controlled to jump to the strong defogging state or the high energy efficiency maintenance state, generating a dual-mode state switching command.

[0030] Specifically, the system reads the pre-stored trigger limit and exit limit benchmarks within the control program. These two benchmark values ​​constitute the basic static hysteresis interval, set to prevent the system from frequently switching near the critical point. For example, based on experience, the trigger limit benchmark is set to a critical safety vector distance of 4, and the exit limit benchmark is set to a critical safety vector distance of 8. Simultaneously, the system reads the trigger limit adjustment range (e.g., [0, 2]) and the exit limit adjustment range (e.g., [0, 3]), and utilizes the fuzzy membership values ​​related to the "dangerous" state generated in the previous step. The two benchmarks are dynamically adjusted using a value range of [0, 1] to determine the reduction magnitude corresponding to the trigger limit benchmark. The calculation formula is: Reduction magnitude = The maximum value of the trigger limit adjustment range, for example, if Then the reduction rate = Determine the expansion range corresponding to the exit limit benchmark, and calculate it using the following formula: Expansion Range = If the maximum value of the exit limit adjustment range is reached, then the expansion range = Based on this, a dynamic short-range trigger limit is generated: Trigger limit baseline - Reduction magnitude = Generate a dynamic long-distance exit limit = Exit limit baseline + Expansion range = Then, the short-distance trigger limit and the long-distance exit limit are compared. When the short-distance trigger limit is not less than the long-distance exit limit in an extreme case, in order to ensure the effectiveness of the hysteresis interval, a preset minimum boundary interval, such as 1.0, must be forcibly maintained. At this time, the short-distance trigger limit is lowered to (new limit sum + new limit difference) / 2 - 0.5, and the long-distance exit limit is raised to (new limit sum + new limit difference) / 2 + 0.5. The combination is completed in the order of short-distance trigger limit first and long-distance exit limit second, forming a double switching boundary.

[0031] Based on the dual switching boundaries, the dynamically calculated short-distance trigger limit (e.g., 2.4) and long-distance exit limit (e.g., 8.6) are extracted. The currently calculated critical safety vector distance is compared with these two limits. The state machine switching logic is as follows: When the critical safety vector distance is less than the short-distance trigger limit (e.g., distance is 2.0, less than 2.4), it indicates a high risk of condensation. A trigger mode switching flag is immediately written, pointing to the strong defogging state, and the continuous dwell counter used to prevent false judgments is simultaneously cleared. When the critical safety vector distance is between the short-distance trigger limit and the long-distance exit limit (e.g., distance is 5.0), this is the hysteresis zone. The system state remains unchanged, whether in the strong defogging or high-efficiency maintenance state, it continues to be maintained, and the continuous dwell counter is cleared. When the critical safety vector distance is greater than the long-distance exit limit, the system state remains unchanged. When the limit is exceeded (e.g., distance is 9.0, greater than 8.6), it indicates that the glass surface is safe and meets the conditions for exiting the high-efficiency defogging mode. At this time, the continuous dwell counter is started and incremented by 1 according to the controller's machine cycle (e.g., 100 milliseconds). A preset number of machine cycles is set as the exit delay, such as 50 cycles (i.e., 5 seconds). This delay is to avoid mode jitter caused by instantaneous fluctuations in environmental parameters. During the continuous dwell count, the current state machine position remains unchanged. When the continuous dwell count exceeds the preset number of 50 cycles, it is confirmed that the exit condition is stably met, and an exit condition met flag is written. This flag points to the high-efficiency maintenance state. Then, according to the written trigger mode switching flag or the exit condition met flag, the state machine transition table is queried to select the corresponding target state, and the state machine is controlled to jump to the high-efficiency defogging state or the high-efficiency maintenance state, generating a dual-mode state switching instruction.

[0032] The specific steps for hardware-based defogging of the driven glass are as follows: Read the dual-mode state switching instruction, extract the state identifier field, switching effect field, instruction cycle field, and state jump target field according to the order of instruction fields, match the state identifier field bit by bit with the strong defogging state code and the high energy efficiency maintenance state code, check whether the switching effect field is in the allowed output state, check whether the state jump target field corresponds to the strong defogging state or the high energy efficiency maintenance state, when the state identifier field matches the strong defogging state code and the state jump target field corresponds to the strong defogging state, register the strong defogging identifier, when the state identifier field matches the high energy efficiency maintenance state code and the state jump target field corresponds to the high energy efficiency maintenance state, register the high energy efficiency maintenance identifier, and obtain the strong defogging identifier or the high energy efficiency maintenance identifier; When registering the high-efficiency defogging flag, the current duty cycle parameter, the target duty cycle parameter for high-efficiency defogging, and the upper limit of the allowable duty cycle for the defogging heating hardware port are read. The target duty cycle parameter for high-efficiency defogging is compared with the upper limit of the allowable duty cycle. The target duty cycle parameter for high-efficiency defogging that does not exceed the upper limit of the allowable duty cycle is selected and written into the high-efficiency defogging output field to generate a large duty cycle control parameter. When registering the high-efficiency maintenance flag, the current duty cycle parameter, the target duty cycle parameter for high-efficiency maintenance, the transition cycle length, and the control point arrangement parameters are read. The current duty cycle parameter is used as the transition start point, and the target duty cycle parameter for high-efficiency maintenance is used as the transition end point. The continuous output positions are divided according to the transition cycle length. The intermediate control points are adjusted according to the change direction of adjacent control points. When the change slope of adjacent output positions is discontinuous, the corresponding intermediate control points are readjusted until the change slope of adjacent output positions is continuous, generating a smooth transition value for the duty cycle. The system reads the pulse width modulation carrier period, timing count upper limit, output effective level, and defogging heating hardware port channel address. It converts the large duty cycle control parameters into effective level duration count values ​​within a single carrier period. The duty cycle smooth transition value is sequentially converted into effective level duration count values ​​for the corresponding carrier period according to the output position within the transition period. The invalid level duration count value for each carrier period is determined based on the timing count upper limit. The effective level duration segment and invalid level duration segment are arranged according to the output effective level. The carrier period, effective level duration count value, invalid level duration count value, and defogging heating hardware port channel address are written into the pulse width modulation output register location to drive the glass defogging hardware to perform defogging.

[0033] Specifically, the dual-mode state switching instruction is read. This instruction is a structured data packet arranged according to a predefined field order. For example, the first two bits of the instruction are the state identifier field, the third bit is the switching activation field, the fourth to seventh bits are the instruction cycle field, and the eighth bit is the state transition target field. These fields are parsed sequentially. The content of the state identifier field is matched bit by bit with the preset strong defogging state code (e.g., binary 01) and high-efficiency maintenance state code (e.g., binary 10) to identify which mode the current instruction belongs to. At the same time, it is checked whether the switching activation field is 1 to confirm that the instruction is valid and allowed to be output. Then, the state transition target field is checked to determine whether it matches the state identifier field. For example, if the status identifier is 01 and the jump target also points to the powerful defogging state, then it is confirmed as a valid mode entry request. When the status identifier field successfully matches the powerful defogging status code "01", and the switch effective field is "1", and the status jump target field also clearly points to the powerful defogging state, the powerful defogging identifier is registered in the internal logic variable. When the status identifier field successfully matches the high energy efficiency maintenance status code "10", and the switch effective field is "1", and the status jump target field also clearly points to the high energy efficiency maintenance state, the high energy efficiency maintenance identifier is registered in the internal logic variable. Through this series of verification and parsing processes, the integrity and correctness of the instruction are ensured, and the powerful defogging identifier or the high energy efficiency maintenance identifier is obtained.

[0034] When the high-efficiency defogging flag is registered, the system enters maximum power output mode. It reads the current pulse width modulation duty cycle parameter, the preset high-efficiency defogging target duty cycle parameter (e.g., 95%), and the upper limit of the allowed duty cycle for the defogging heating hardware port stored in the hardware configuration (e.g., 98%, set for hardware protection). It compares the high-efficiency defogging target duty cycle parameter 95% and the upper limit of the allowed duty cycle 98%, selecting the smaller value as the final output value, i.e., 95%. This value is written to the internal high-efficiency defogging output field, generating a high duty cycle control parameter. When the high-efficiency maintenance flag is registered, the system smoothly exits from high-efficiency defogging mode and reads the current duty cycle parameter. The parameters include (e.g., 95%), a preset high-efficiency maintenance target duty cycle parameter (e.g., 30%), a transition period length (e.g., 10 seconds), and control point arrangement parameters. These parameters define the shape of the transition curve, such as using a Bézier curve or an S-curve. The transition curve starts at the current duty cycle of 95% and ends at the high-efficiency maintenance target duty cycle of 30%. Based on a 10-second transition period and a 100-millisecond control cycle, the transition process is divided into 100 consecutive output positions. The duty cycle for each output position is calculated using the control point arrangement parameters, for example, by interpolating using a cubic Bézier curve. The formula is... ,in Normalized time from 0 to 1 Starting from 95%, 30% of the endpoint, control point and Based on the curve shape requirements, the slope of the duty cycle calculated from adjacent output positions is checked. If an abrupt change or discontinuity in the slope is detected, the control points are fine-tuned. and The interpolation is recalculated at the position until the slope of the duty cycle change is continuous throughout the entire transition period, generating a series of duty cycle values ​​that change smoothly over time, which are the duty cycle smooth transition values.

[0035] Read the relevant configuration parameters of the pulse width modulation hardware timer, including the carrier period (e.g., 100 microseconds, corresponding to a 10kHz frequency), the timer count limit (e.g., 1000, which determines the duty cycle resolution), the output active level (high or low active), and the channel address of the target defogging heating hardware port. For large duty cycle control parameters, such as 95%, convert them into an active level duration count value within a single carrier period. The calculation method is: Active level count value = Timer count limit × Duty cycle = For the duty cycle smooth transition value, which is a duty cycle sequence, each duty cycle value (e.g., 94.5%, 94.0%, ... in the sequence) needs to be converted into an effective level duration count value (i.e., 945, 940, ...) for the corresponding carrier cycle according to its output position within the transition period. Based on the timing count upper limit of 1000 and the calculated effective level duration count value of 950, the ineffective level duration count value for each carrier cycle is determined. According to the preset output valid level (e.g., high level valid), the timer is configured to output a high level from count 0 to 949 and a low level from count 950 to 999, forming a complete PWM waveform. Then, the calculated carrier period configuration, valid level continuous count value, invalid level continuous count value, and target hardware port channel address are written into the microcontroller's pulse width modulation output register group. The hardware automatically generates a continuous PWM waveform to drive the external power switching device, thereby controlling the current flowing through the heating wire and driving the glass defogging hardware to perform defogging.

Claims

1. A smart glass defogging system based on dew point dynamic tracking and dual-mode collaborative control, characterized in that, The system includes: The environmental parameter discrete module collects temperature and relative humidity parameters from physical sensors and reads pulse width modulation heat integration parameters from the defogging heating hardware port. It constructs a two-dimensional thermal grid coordinate system using the physical sensor temperature parameters, the pulse width modulation heat integration parameters, and preset glass heat capacity and area size parameters, and calculates and generates a set of grid node temperatures. It extracts the calculated temperature corresponding to the lowest value node within the grid node temperature set and combines it with the relative humidity parameter to generate extreme value node state coordinates. The enthalpy-humidity matrix addressing module inputs the extreme node state coordinates into a multidimensional digital enthalpy-humidity matrix to generate a thermodynamic state mapping value; it calculates the spatial vector difference between the thermodynamic state mapping value and the dew point critical curve parameters pre-stored in the control program to generate a critical safety vector distance. The fuzzy inference debouncing module inputs the critical safety vector distance into the fuzzy logic program layer to generate fuzzy membership degree values; generates short-distance trigger limits and long-distance exit limits based on the fuzzy membership degree values, and merges them to construct a dual switching boundary; and generates a dual-mode state switching command based on the comparison result between the critical safety vector distance and the dual switching boundary values. The dual-mode collaborative output module reads the dual-mode state switching instruction. When a high-efficiency defogging indicator is obtained, it generates a large duty cycle control parameter. When a high-efficiency maintenance indicator is obtained, it calculates and generates a duty cycle smooth transition value based on the current duty cycle parameter. The large duty cycle control parameter or the duty cycle smooth transition value is converted into pulse width modulation waveform data to drive the glass defogging hardware to defog.

2. The intelligent glass defogging system based on dew point dynamic tracking and dual-mode collaborative control according to claim 1, characterized in that, The steps for obtaining the temperature set of the grid nodes are as follows: Within the same sampling period, the temperature and relative humidity parameters of the physical sensor are collected. The time correspondence between the temperature and relative humidity parameters of each group of physical sensor is recorded according to the sampling time. The pulse width modulation heat integration parameters of the defogging heating hardware port within the same sampling period are read. The preset glass heat capacity parameters and area size parameters are called. The grid row boundary, grid column boundary, number of nodes and node spatial position are determined according to the area size parameters. The heat capacity attribute is configured for each node according to the glass heat capacity parameters. The physical sensor temperature parameters are written to the glass boundary nodes, and the pulse width modulation heat integration parameters are written to the boundary nodes corresponding to the heating port to form a two-dimensional thermal grid coordinate. Based on the coordinates of the two-dimensional thermal grid, the row and column positions, heat capacity attributes, boundary temperatures, and heating boundary heat of each node are read row by row. According to the spatial position of the nodes, the adjacent nodes of each internal node in the row front, row back, column front, and column back are determined. The difference between the estimated temperature of each adjacent node and the estimated temperature of the current node is converted into row-direction heat change and column-direction heat change according to the node spatial interval. The estimated temperature of the current node is updated by combining the heat capacity attribute of the current node and the heating boundary heat of the current iteration cycle. After completing one round of updates, the estimated temperature difference of nodes in two adjacent rounds is compared one by one. When the estimated temperature difference of all nodes does not exceed the preset convergence limit, the update stops. When the estimated temperature difference of any node exceeds the preset convergence limit, the next round of updates continues. The estimated temperatures of each node when the update stops are summarized to generate a set of grid node temperatures.

3. The intelligent glass defogging system based on dew point dynamic tracking and dual-mode collaborative control according to claim 1, characterized in that, The steps for obtaining the state coordinates of the extreme node are as follows: Based on the set of grid node temperatures, the estimated temperatures of each node are read sequentially according to a fixed row and column order. The estimated temperature of the first node is registered as the current minimum temperature. The estimated temperatures of subsequent nodes are compared with the current minimum temperature. When the estimated temperature of a subsequent node is less than the current minimum temperature, the current minimum temperature is updated and the corresponding node is registered. When the estimated temperature of a subsequent node is not less than the current minimum temperature, the current minimum temperature and the corresponding node remain unchanged. After traversing all nodes, the node with the lowest value is determined. The estimated temperature corresponding to the node with the lowest value and the relative humidity parameter at the same sampling time are extracted. The estimated temperature is converted into a matrix row index value according to a preset temperature index interval, and the relative humidity parameter is converted into a matrix column index value according to a preset humidity index interval. When the estimated temperature or relative humidity parameter is located at the boundary of an adjacent index interval, the absolute difference between the parameter value and the center value of the adjacent interval is compared. The interval number with the smallest absolute difference is selected as the corresponding index value. The matrix row index value and the matrix column index value are combined to generate the state coordinates of the extreme value node.

4. The intelligent glass defogging system based on dew point dynamic tracking and dual-mode collaborative control according to claim 1, characterized in that, The steps for obtaining the thermodynamic state mapping value are as follows: The system invokes a pre-stored multidimensional digital enthalpy-humidity matrix in the microprocessor memory, reads the row index range, column index range, dimensional arrangement order, and intersection storage location corresponding to the multidimensional digital enthalpy-humidity matrix, sequentially extracts the matrix row index value and matrix column index value from the extreme value node state coordinates, matches the matrix row index value with the row index range of the multidimensional digital enthalpy-humidity matrix item by item, and matches the matrix column index value with the column index range of the multidimensional digital enthalpy-humidity matrix item by item. When both the matrix row index value and the matrix column index value fall within the corresponding index range, the row and column intersection point pointed to by the matrix row index value and the matrix column index value are locked, the thermodynamic data stored in the row and column intersection point according to the dimensional arrangement order is read, and the thermodynamic data is combined according to the original dimensional order to generate a thermodynamic state mapping value.

5. The intelligent glass defogging system based on dew point dynamic tracking and dual-mode collaborative control according to claim 1, characterized in that, The steps for obtaining the critical safety vector distance are as follows: The system reads the pre-stored dew point critical curve parameters from the control program, parses the curve node sequence, state dimension identifier, and component values ​​of each curve node corresponding to the dew point critical curve parameters, expands the thermodynamic state mapping value into a state component sequence according to the state dimension identifier, arranges the components of each curve node in the dew point critical curve parameters according to the same state dimension identifier, compares the state component sequence of the thermodynamic state mapping value with the component arrangement order of each curve node, excludes curve nodes with inconsistent state dimension identifiers, calculates the component value difference of each dimension for curve nodes with consistent state dimension identifiers, records the component value difference and positive and negative directions of each component according to the curve node sequence, and generates a spatial vector difference value. Based on the spatial vector difference, the component value difference corresponding to each curve node is read in groups. The component value difference of each dimension under the same curve node is squared. All squared results under the same curve node are accumulated. The accumulated results are square rooted to obtain the vector distance value corresponding to each curve node. All vector distance values ​​are compared in the order of curve nodes of the dew point critical curve parameter. The first vector distance value is registered as the current minimum distance. When the subsequent vector distance value is less than the current minimum distance, the current minimum distance is updated. When the subsequent vector distance value is not less than the current minimum distance, the current minimum distance is kept unchanged. After completing the comparison of all vector distance values, the critical safe vector distance is generated.

6. The intelligent glass defogging system based on dew point dynamic tracking and dual-mode collaborative control according to claim 1, characterized in that, The steps for obtaining the fuzzy membership degree value are as follows: The critical safety vector distance is input into the fuzzy logic program layer. The pre-stored distance segment boundaries, membership interval endpoints, interval increasing direction, and interval decreasing direction within the fuzzy logic program layer are read. The corresponding distance segments are retrieved according to the numerical position of the critical safety vector distance. The membership interval endpoints corresponding to both ends of the distance segments are extracted. The interval position ratio of the critical safety vector distance relative to the starting boundary of the distance segment is calculated. The numerical proportion of the membership interval endpoints is adjusted according to the interval position ratio. When the critical safety vector distance is located at the common boundary of adjacent distance segments, the membership values ​​corresponding to the adjacent distance segments are calculated respectively. The membership value with the largest value is selected to generate the fuzzy membership degree value.

7. The intelligent glass defogging system based on dew point dynamic tracking and dual-mode collaborative control according to claim 1, characterized in that, The steps for obtaining the dual-mode state switching command are as follows: The system reads the pre-stored trigger limit benchmark, exit limit benchmark, trigger limit adjustment range, and exit limit adjustment range within the control program. Based on the numerical position of the fuzzy membership degree value in the membership interval, it determines the reduction range corresponding to the trigger limit benchmark and the expansion range corresponding to the exit limit benchmark. It subtracts the reduction range from the trigger limit benchmark to generate a short-distance trigger limit. It adds the expansion range to the exit limit benchmark to generate a long-distance exit limit. It compares the short-distance trigger limit and the long-distance exit limit. When the short-distance trigger limit is not less than the long-distance exit limit, it lowers the short-distance trigger limit according to a preset boundary interval and raises the long-distance exit limit according to the same preset boundary interval. It completes the combination in the order of short-distance trigger limit first and long-distance exit limit second, forming a dual switching boundary. Based on the aforementioned dual switching boundaries, short-distance trigger limits and long-distance exit limits are extracted. The critical safety vector distance is compared numerically with both the short-distance trigger limit and the long-distance exit limit. When the critical safety vector distance is less than the short-distance trigger limit, a trigger mode switching flag is written, and the continuous dwell count is cleared. When the critical safety vector distance is between the short-distance trigger limit and the long-distance exit limit, the current state machine position is maintained, and the continuous dwell count is cleared. When the critical safety vector distance is greater than the long-distance exit limit, the continuous dwell count is incremented sequentially according to machine cycles. When the continuous dwell count does not exceed the preset number of machine cycles, the current state machine position is maintained. When the continuous dwell count exceeds the preset number of machine cycles, an exit condition fulfillment flag is written. The corresponding state machine target position is selected according to the trigger mode switching flag or the exit condition fulfillment flag, and the state machine is controlled to jump to either the high-efficiency defogging state or the high-efficiency maintenance state, generating a dual-mode state switching command.

8. The intelligent glass defogging system based on dew point dynamic tracking and dual-mode collaborative control according to claim 1, characterized in that, The specific steps for hardware-based defogging of the driven glass are as follows: Read the dual-mode state switching instruction, extract the state identifier field, switching effect field, instruction cycle field, and state jump target field according to the order of instruction fields, match the state identifier field bit by bit with the strong defogging state code and the high energy efficiency maintenance state code, check whether the switching effect field is in the allowed output state, check whether the state jump target field corresponds to the strong defogging state or the high energy efficiency maintenance state, when the state identifier field matches the strong defogging state code and the state jump target field corresponds to the strong defogging state, register the strong defogging identifier, when the state identifier field matches the high energy efficiency maintenance state code and the state jump target field corresponds to the high energy efficiency maintenance state, register the high energy efficiency maintenance identifier, and obtain the strong defogging identifier or the high energy efficiency maintenance identifier; When registering the powerful defogging identifier, read the current duty cycle parameter, the target duty cycle parameter of powerful defogging, and the upper limit of the allowed duty cycle of the defogging heating hardware port. Compare the target duty cycle parameter of powerful defogging with the upper limit of the allowed duty cycle. Select the target duty cycle parameter of powerful defogging that does not exceed the upper limit of the allowed duty cycle and write it into the powerful defogging output field to generate the large duty cycle control parameter. When registering the high-efficiency maintenance flag, the current duty cycle parameter, the high-efficiency maintenance target duty cycle parameter, the transition period length, and the control point arrangement parameters are read. The current duty cycle parameter is used as the transition start point, and the high-efficiency maintenance target duty cycle parameter is used as the transition end point. The continuous output positions are divided according to the transition period length. The intermediate control points are adjusted according to the change direction of adjacent control points. When the change slope of adjacent output positions is discontinuous, the corresponding intermediate control points are readjusted until the change slope of adjacent output positions is continuous, and a smooth transition value of the duty cycle is generated. The pulse width modulation carrier period, timing count upper limit, output effective level, and defogging heating hardware port channel address are read. The large duty cycle control parameters are converted into effective level duration count values ​​within a single carrier period. The duty cycle smooth transition value is sequentially converted into effective level duration count values ​​for the corresponding carrier period according to the output position within the transition period. The ineffective level duration count value for each carrier period is determined according to the timing count upper limit. The effective level duration segment and ineffective level duration segment are arranged according to the output effective level. The carrier period, effective level duration count value, ineffective level duration count value, and defogging heating hardware port channel address are written into the pulse width modulation output register location to drive the glass defogging hardware to perform defogging.

Images