Intelligent control system for kitchen island environment based on IoT and sensor data

Abstract

This invention discloses an intelligent control system for the kitchen island environment based on the Internet of Things (IoT) and sensor data, belonging to the field of intelligent furniture environment control technology. It includes: a data preprocessing module for generating a kitchen island environment dataset; an island zoning modeling module for constructing an island zoning model; a linkage diagram construction module for constructing an island environment linkage diagram; a scene state recognition module for generating zoning scene states; a control constraint generation module for generating a control constraint set; an optimization decision module for generating zoning linkage control strategies using the NSGA-III multi-objective optimization algorithm; and a feedback correction module for generating intelligent control results for the kitchen island environment. This invention achieves zoning perception, scene recognition, and linkage control of an open kitchen island area, possessing advantages such as high control accuracy, strong scene adaptability, and good device synergy.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent furniture environment regulation, and particularly to an intelligent regulation system for the kitchen island environment based on the Internet of Things and sensor data. Background Art

[0002] In recent years, with the development of smart home, Internet of Things communication, sensor perception, and kitchen appliance control technologies, open kitchens have gradually transformed from a single cooking space into a composite space for meal preparation, cooking, dining, and communication. Existing kitchen environment regulation technologies usually collect kitchen environment data through temperature and humidity sensors, oil fume sensors, gas sensors, human body sensing devices, and intelligent appliance controllers, and control lighting devices, ventilation devices, air purification devices, temperature and humidity regulation devices, and safety reminder devices according to fixed thresholds or user instructions to achieve functions such as oil fume emission, air purification, lighting adjustment, and safety alarm. Some solutions also access multiple kitchen devices through a smart gateway or cloud platform, enabling different devices to perform linkage control according to preset scenarios.

[0003] However, the existing open kitchen environment regulation methods are still difficult to meet the localized use requirements of the kitchen island area. On the one hand, existing solutions mostly take the entire kitchen space or a single appliance as the control object, lacking partition modeling for the island boundary, connection form, and passage position, and unable to distinguish the environmental change differences in the operation partition, passage partition, and diffusion partition. On the other hand, there is a lack of graph-structured organization between the sensor coverage range and the device action range, and it is difficult for environmental perception data, personnel activity data, and device operation data to form a partition-oriented fusion judgment. Existing control methods often rely on fixed rules and lack multi-objective collaborative decision-making on environmental improvement, energy consumption, noise, and safety priorities, resulting in insufficient environmental regulation accuracy and scenario adaptability in the island area. Summary of the Invention

[0004] An object of the present invention is to propose an intelligent regulation system for the kitchen island environment based on the Internet of Things and sensor data. The present invention makes full use of Internet of Things sensor perception, island partition modeling, environmental linkage graph, and NSGA-III multi-objective optimization decision-making technology, and details the implementation process of environmental state recognition, partition linkage regulation, and feedback correction in the open kitchen island area, with the advantages of accurate local perception, strong scenario adaptability, high device cooperation degree, and sustainable optimization of regulation effects.

[0005] According to an embodiment of the present invention, an intelligent regulation system for the kitchen island environment based on the Internet of Things and sensor data includes:

[0006] A data preprocessing module, configured to collect kitchen island regulation-related data and perform preprocessing to generate a kitchen island environment data set;

[0007] The island partitioning modeling module is used to divide the operation partitioning, passage partitioning, and diffusion partitioning according to the island boundary, connection form, and passage location, and to construct the island partitioning model.

[0008] The linkage diagram construction module is used to construct an island environment linkage diagram based on the island zoning model, which associates the sensor coverage area and the equipment operating range.

[0009] The scene state recognition module is used to generate the scene state of each zone by integrating the kitchen island environment dataset with the island environment linkage diagram.

[0010] The regulation constraint generation module is used to determine environmental improvement goals, energy consumption constraints, noise constraints, and safety priorities based on the partitioned scenario state, and generate a set of regulation constraints;

[0011] The optimization decision module is used to perform non-dominated sorting and reference point screening of the set of control constraints using the NSGA-III multi-objective optimization algorithm, determine the master control equipment, coordinating equipment, control amplitude and duration, and generate a zone linkage control strategy.

[0012] The feedback correction module is used to generate equipment scheduling instructions based on the zone linkage control strategy, and correct the island environment linkage diagram based on the regulated zone scene status, thereby generating intelligent control results for the kitchen island environment.

[0013] Optionally, the kitchen island control-related data includes environmental perception data, personnel activity data, kitchen safety perception data, environmental control equipment operation data, and IoT device communication status data. The preprocessing includes data collection time alignment, device identification unification, data format conversion, abnormal data collection value removal, missing data collection value completion, data unit unification, and partition label writing processing.

[0014] Optionally, the island partitioning modeling module includes:

[0015] Retrieve island boundary data, connection shape data, and passage location data from the kitchen island environment dataset;

[0016] Map the island boundary data to the kitchen plane coordinate system to generate the island boundary coordinates;

[0017] Based on the connection morphology data, the connecting side, open side, and adjacent side in the island platform boundary coordinates are marked to generate island platform boundary marking results;

[0018] Based on the passage location data, the movement paths and stopping locations of people around the island platform are marked, and passage location marking results are generated;

[0019] Based on the island / platform boundary marking results and access location marking results, the operation zone, access zone, and diffusion zone are divided;

[0020] Generate a partition adjacency matrix based on the adjacency between the operation partition, the passage partition, and the diffusion partition;

[0021] The island boundary coordinates, island boundary marking results, passage position marking results, and partition adjacency matrix are combined to construct the island partition model.

[0022] Optionally, the linkage diagram construction module includes:

[0023] Based on the sensor installation location, sensing direction, and sensing distance, the sensor coverage area is matched with the boundaries of each zone to determine the sensor coverage zone;

[0024] Based on the installation location, direction of action, and distance of action of the kitchen environment control equipment, the range of action of the equipment is matched with the boundaries of each zone to determine the zones in which the equipment operates;

[0025] Write the operation zone, passage zone, and diffusion zone as zone nodes, write the sensors as sensing nodes, and write the kitchen environment control equipment as device nodes.

[0026] Based on the sensor coverage partition, a sensing coverage edge is written between the sensing node and the partition node. The sensing coverage edge includes the main sensing coverage edge and the auxiliary sensing coverage edge.

[0027] Based on the device function partition, write device function edges between the device node and the partition node. Device function edges include master device function edges and cooperating device function edges.

[0028] Write partition adjacency edges between adjacent partition nodes based on the partition adjacency matrix;

[0029] By combining partition nodes, sensing nodes, device nodes, sensing coverage edges, device action edges, and partition adjacency edges, an island-level environmental linkage diagram is constructed.

[0030] Optionally, the scene state recognition module includes:

[0031] The system retrieves partition nodes, sensing nodes, equipment nodes, sensing coverage edges, equipment action edges, and partition adjacency edges from the island environment linkage diagram, and retrieves sensor data, equipment operation data, and personnel activity data from the kitchen island environment dataset.

[0032] Based on the sensing coverage edge, sensor data is mapped to the corresponding partition node to generate partition environmental features;

[0033] Based on the device action edge, the device operation data is mapped to the corresponding partition node to generate partition device characteristics;

[0034] Generate zoning personnel characteristics based on personnel activity data;

[0035] Based on the partition adjacency edges, the partition environment features, partition device features, and partition personnel features of adjacent partition nodes are fused to generate partition fusion features;

[0036] Based on the partition fusion characteristics, determine the food preparation status, cooking status, cleaning status, dining status, unattended status, and safety anomaly status of each partition node, and generate the partition scene status.

[0037] Optionally, the regulation constraint generation module includes:

[0038] Read the partition scene status according to the partition number;

[0039] Determine the environmental improvement goals for the corresponding partition based on the status markers in the partition scenario status;

[0040] Energy consumption and noise constraints for the corresponding zones are determined based on environmental improvement goals;

[0041] Determine the security priority of the corresponding partition based on the status markers and environmental improvement goals in the partition scenario state;

[0042] The environmental improvement goals, energy consumption constraints, noise constraints, and safety priorities are combined according to the zoning number to generate a set of control constraints.

[0043] Optionally, the optimization decision module includes:

[0044] Read the set of control constraints and generate multiple candidate control schemes based on the set of control constraints. Each candidate control scheme includes the master control device, the coordinating device, the control amplitude, and the duration.

[0045] The degree to which each candidate control scheme meets the environmental improvement target, energy consumption constraint, noise constraint, and safety priority is satisfied is used as the optimization objective.

[0046] The NSGA-III multi-objective optimization algorithm is used to perform non-dominated ranking of multiple candidate control schemes to generate the dominance level of the candidate control schemes.

[0047] Based on the reference point screening mechanism of the NSGA-III multi-objective optimization algorithm, reference point association and crowding screening are performed on candidate control schemes within the same dominance level to determine the target control scheme.

[0048] The main control equipment, coordinating equipment, control amplitude, and duration in the target control scheme are combined according to the partition number to generate a partition linkage control strategy.

[0049] Optionally, the feedback correction module includes:

[0050] Extract the master control device, cooperating devices, control amplitude and duration from the zone linkage control strategy, arrange the execution order according to safety priority, and generate equipment scheduling instructions;

[0051] Send the equipment scheduling command to the corresponding kitchen environment control equipment for execution, and obtain the controlled zoning scene status after execution;

[0052] The state of the partitioned scenario after adjustment is compared with the state of the partitioned scenario before execution to determine the state change results of each partition node.

[0053] The device action edges and partition adjacency edges in the island environment linkage diagram are corrected based on the state change results;

[0054] The system combines equipment scheduling instructions, adjusted zoning scene status, and corrected island environment linkage diagram to generate intelligent control results for the kitchen island environment.

[0055] The beneficial effects of this invention are:

[0056] This invention generates a kitchen island environment dataset by collecting and preprocessing kitchen island control-related data. The island zoning modeling module divides the island into operation zones, passage zones, and diffusion zones based on island boundaries, connection patterns, and access locations, and constructs an island zoning model. This transforms the island area in an open kitchen from being treated as a typical kitchen space into a zoning representation tailored to specific usage conditions, thus improving the accuracy of island area environmental control.

[0057] This invention uses a linkage graph construction module to associate sensor coverage and equipment operating range, constructing an island environment linkage graph that includes partition nodes, sensing nodes, equipment nodes, and related edges. This allows sensor perception results, equipment operating range, and partition adjacency to form a unified graph structure. A scene state recognition module, based on the island environment linkage graph and fused with a kitchen island environment dataset, generates partitioned scene states. This enables separate identification of food preparation, cooking, cleaning, dining, unattended, and safety anomaly states for different partitions, avoiding rough judgments based solely on a single sensor or the average state of the entire kitchen.

[0058] This invention uses a constraint generation module to determine environmental improvement goals, energy consumption constraints, noise constraints, and safety priorities based on the state of the partitioned scene. An optimization decision module employs the NSGA-III multi-objective optimization algorithm for non-dominated sorting and reference point selection to determine the master control device, cooperating devices, control amplitude, and duration. This enables the kitchen environment control equipment to make coordinated decisions among environmental improvement, energy consumption control, noise limitation, and safety response. A feedback correction module modifies the island's environmental linkage diagram based on the controlled partitioned scene state, allowing the device action edges and partition adjacency edges to update with the actual control effect, improving the adaptability and continuous optimization capability of subsequent partitioned linkage control strategies. Attached Figure Description

[0059] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0060] Figure 1 This is an overall flowchart of the intelligent control system for the kitchen island environment based on the Internet of Things and sensor data proposed in this invention;

[0061] Figure 2 This is a schematic diagram of the island environment linkage diagram of the intelligent control system for kitchen island environment based on Internet of Things and sensor data proposed in this invention.

[0062] Figure 3 This is a schematic diagram illustrating the zoned linkage control strategy of the intelligent kitchen island environment control system based on the Internet of Things and sensor data proposed in this invention. Detailed Implementation

[0063] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, illustrating only the basic structure of the invention, and therefore only show the components relevant to the invention.

[0064] refer to Figures 1-3 A smart kitchen island environment control system based on the Internet of Things and sensor data includes:

[0065] The data preprocessing module is used to collect and preprocess kitchen island control-related data to generate a kitchen island environment dataset.

[0066] The island partitioning modeling module is used to divide the operation partitioning, passage partitioning, and diffusion partitioning according to the island boundary, connection form, and passage location, and to construct the island partitioning model.

[0067] The linkage diagram construction module is used to construct an island environment linkage diagram based on the island zoning model, which associates the sensor coverage area and the equipment operating range.

[0068] The scene state recognition module is used to generate the scene state of each zone by integrating the kitchen island environment dataset with the island environment linkage diagram.

[0069] The regulation constraint generation module is used to determine environmental improvement goals, energy consumption constraints, noise constraints, and safety priorities based on the partitioned scenario state, and generate a set of regulation constraints;

[0070] The optimization decision module is used to perform non-dominated sorting and reference point screening of the set of control constraints using the NSGA-III multi-objective optimization algorithm, determine the master control equipment, coordinating equipment, control amplitude and duration, and generate a zone linkage control strategy.

[0071] The feedback correction module is used to generate equipment scheduling instructions based on the zone linkage control strategy, and correct the island environment linkage diagram based on the regulated zone scene status, thereby generating intelligent control results for the kitchen island environment.

[0072] In this embodiment, the kitchen island control-related data includes environmental perception data, personnel activity data, kitchen safety perception data, environmental control equipment operation data, and IoT device communication status data. Preprocessing includes data collection time alignment, device identification unification, data format conversion, abnormal data collection value removal, missing data collection value completion, data unit unification, and partition label writing.

[0073] In this embodiment, the island platform partitioning modeling module includes:

[0074] Retrieve island boundary data, connection shape data, and passage location data from the kitchen island environment dataset;

[0075] Map the island boundary data to the kitchen plane coordinate system to generate the island boundary coordinates;

[0076] The generation of island boundary coordinates is as follows: Obtain the countertop outline points, side length data, corner data, and boundary distance data from the island boundary data. Using the kitchen floor projection as the coordinate plane and the fixed kitchen reference point as the coordinate origin, establish a horizontal coordinate axis along the direction of the cabinet reference edge and a vertical coordinate axis perpendicular to the cabinet reference edge and pointing towards the open area of ​​the kitchen, forming a kitchen planar coordinate system. Based on the horizontal and vertical distances of the countertop outline points relative to the coordinate origin, convert each countertop outline point into planar point coordinates, generating a set of outline point coordinates. Verify the connection distance between adjacent outline points based on the side length data, verify the corner direction between adjacent boundary line segments based on the corner data, and correct any deviations in the outline point coordinates based on the boundary distance data. Connect the outline point coordinate sets according to the arrangement order of the countertop outline points on the island's outer contour to generate a closed boundary coordinate sequence, which is then used as the island boundary coordinates.

[0077] Based on the connection pattern data, the connecting side, open side, and adjacent side in the island boundary coordinates are marked to generate island boundary marking results. Among them, the connecting side refers to the side of the island boundary that is directly connected to the fixed kitchen structure, the open side refers to the side of the island boundary that faces the open kitchen space, is not connected to the wall or cabinet, and is accessible or accessible to people, and the adjacent side refers to the side of the island boundary that is not directly connected to the wall or cabinet, but is close to the adjacent functional area or equipment area.

[0078] Based on the passage location data, the movement paths and stopping locations of people around the island platform are marked, and passage location marking results are generated;

[0079] Based on the island's boundary markings and access location markings, the kitchen is divided into operation zones, access zones, and diffusion zones. Specifically, the positions of the connecting side, open side, and adjacent side in the kitchen's planar coordinate system are determined based on the island's boundary markings, and the positions of personnel movement paths and resting positions in the kitchen's planar coordinate system are determined based on the access location markings. The area surrounding the island that is adjacent to the open side and adjacent side, and coincides with or intersects with the resting positions, is marked as a candidate operation zone; the area that coincides with or intersects with the personnel movement paths and is located outside the island's outer contour is marked as a candidate access zone. Areas located outside the open side, outside the adjacent side, or outside the projected area above the island platform, and extending along the direction of the spread of fumes, heat, moisture, or odors, are marked as candidate diffusion areas. The overlapping portions of candidate operation areas, candidate passage areas, and candidate diffusion areas are assigned to different zones: those coinciding with the stopping position are assigned to the operation zone; those coinciding with the personnel movement path are assigned to the passage zone; and those coinciding with the direction of spread are assigned to the diffusion zone. After assignment determination, the boundaries of each zone are merged and zoned, generating operation zones, passage zones, and diffusion zones.

[0080] Generate a partition adjacency matrix based on the adjacency between the operation partition, the passage partition, and the diffusion partition;

[0081] The generation of the partition adjacency matrix is ​​as follows: Partition numbers are written to the operation partition, passage partition, and diffusion partition; the boundary coordinates of each partition are converted to a unified coordinate format; the adjacency determination distance is calculated based on the platform width in the island boundary data and the personnel movement path width in the passage location data, and the smaller value between the platform width and the personnel movement path width is used as the adjacency determination distance; any two partitions are selected one by one, and it is determined whether the boundaries of the two partitions overlap, intersect, or the distance between them is less than the adjacency determination distance; when the boundaries of two partitions overlap, intersect, or the distance between them is less than the adjacency determination distance, an adjacency marker is written between the corresponding two partition numbers; when the boundaries of two partitions do not overlap, intersect, and the distance between them is not less than the adjacency determination distance, a non-adjacency marker is written between the corresponding two partition numbers; a matrix row and column index is established according to the order of all partition numbers, and the adjacency marker and non-adjacency marker are written to the corresponding matrix positions; the local marker is written to the diagonal position of the matrix, and the same adjacency marker is written to two adjacent matrix positions to generate the partition adjacency matrix;

[0082] The island boundary coordinates, island boundary marking results, passage position marking results, and partition adjacency matrix are combined to construct the island partition model.

[0083] In this embodiment, the linkage diagram construction module includes:

[0084] Based on the sensor installation location, sensing direction, and sensing distance, the sensor coverage area is matched with the boundaries of each zone to determine the sensor coverage zone;

[0085] The determination of sensor coverage zones is as follows: The sensor installation locations are transformed to the kitchen plane coordinate system to generate sensor position coordinates; starting from the sensor position coordinates, the sensor center sensing line is marked along the sensing direction, and the ending position of the center sensing line is determined according to the sensing distance; the corresponding coverage pattern is called according to the sensor type to generate the sensor coverage area, where fixed-point sensors form a circular coverage area according to the sensing distance, directional sensors form a fan-shaped coverage area according to the sensing direction and sensing distance, and linear sensors form a strip-shaped coverage area according to the sensing direction and sensing distance; the sensor coverage area is overlapped and matched with the boundaries of the operation zone, passage zone, and diffusion zone, and the coverage portion of the sensor coverage area falling within the boundaries of each zone is counted; the zone with the largest coverage area is determined as the main sensor coverage zone, and the zones with coverage portions but not determined as the main sensor coverage zone are determined as auxiliary sensor coverage zones; the main sensor coverage zone and the auxiliary sensor coverage zone are combined to determine the sensor coverage zone.

[0086] Based on the installation location, direction of action, and distance of action of the kitchen environment control equipment, the effective range of the equipment is matched with the boundaries of each zone to determine the effective zones of the equipment; among which, the kitchen environment control equipment includes lighting equipment, ventilation equipment, air purification equipment, temperature and humidity control equipment, and safety reminder equipment;

[0087] The determination of equipment function zones is as follows: The installation location is converted to the kitchen plane coordinate system to generate equipment location coordinates; the function pattern of the equipment is determined according to its type, where lighting equipment forms a lighting function zone based on its illumination direction and distance, ventilation equipment forms a ventilation function zone based on its intake and exhaust direction and distance, air purification equipment forms a purification function zone based on its air supply direction and distance, temperature and humidity control equipment forms a temperature and humidity function zone based on its air supply direction and distance, and safety reminder equipment forms a reminder function zone based on its reminder coverage direction and distance; the equipment function zones are overlapped and matched with the boundaries of the operation zone, passage zone, and diffusion zone, and the effective portion of the equipment function zone falling within each zone boundary is counted; the zone with the largest effective area is determined as the primary function zone, and zones with effective portions but not determined as primary function zones are determined as auxiliary function zones; the primary function zones and auxiliary function zones are combined to determine the equipment function zones.

[0088] Write the operation zone, passage zone, and diffusion zone as zone nodes, write the sensors as sensing nodes, and write the kitchen environment control equipment as device nodes.

[0089] Based on the sensor coverage partition, a sensing coverage edge is written between the sensing node and the partition node. The sensing coverage edge includes the main sensing coverage edge and the auxiliary sensing coverage edge.

[0090] Based on the device function partition, write device function edges between the device node and the partition node. Device function edges include master device function edges and cooperating device function edges.

[0091] Write partition adjacency edges between adjacent partition nodes based on the partition adjacency matrix;

[0092] By combining partition nodes, sensing nodes, device nodes, sensing coverage edges, device action edges, and partition adjacency edges, an island-level environmental linkage diagram is constructed.

[0093] In this embodiment, the scene state recognition module includes:

[0094] The system retrieves partition nodes, sensing nodes, equipment nodes, sensing coverage edges, equipment action edges, and partition adjacency edges from the island environment linkage diagram, and retrieves sensor data, equipment operation data, and personnel activity data from the kitchen island environment dataset.

[0095] Based on the sensing coverage edge, sensor data is mapped to the corresponding partition node to generate partition environmental features;

[0096] The generation of partitioned environmental features is as follows: according to the sensing coverage edge, read the connection status between each sensing node and the partition node, and write the sensor data corresponding to the sensing node into the connected partition node; when a partition node is connected to multiple sensing nodes, organize the sensor data according to the order of priority of the main sensing coverage edge over the auxiliary sensing coverage edge, and normalize the sensor data at the same acquisition time; arrange the normalized sensor data according to the partition number and acquisition time to generate partitioned environmental features.

[0097] Based on the device action edge, the device operation data is mapped to the corresponding partition node to generate partition device characteristics;

[0098] The generation of partition device features is as follows: read the connection status between each device node and the partition node according to the device action edge, and write the device operation data corresponding to the device node into the connected partition node; when a partition node is connected to multiple device nodes, organize the device operation data according to the order of the main device action edge taking precedence over the cooperating device action edge, and convert the device start / stop status, operating level, and operating duration into a unified status code; arrange the unified status codes according to the partition number and the collection time to generate partition device features;

[0099] Generate zoning personnel characteristics based on personnel activity data;

[0100] The generation of personnel characteristics by partition is as follows: matching personnel activity data to the corresponding partition nodes according to personnel location, dwell time and movement direction in personnel activity data; encoding the existence status, dwell status and movement status of personnel at the same partition node at the same collection time to generate partition personnel status codes; arranging the partition personnel status codes according to partition number and collection time to generate partition personnel characteristics;

[0101] Based on the partition adjacency edges, the partition environment features, partition device features, and partition personnel features of adjacent partition nodes are fused to generate partition fusion features;

[0102] The generation of partition fusion features is as follows: A partition node is selected as the current partition node. Adjacent partition nodes directly connected to the current partition node are queried based on their adjacent edges. The partition environment features, partition equipment features, and partition personnel features of the current partition node are organized according to the acquisition time. The partition environment features, partition equipment features, and partition personnel features at the same acquisition time are concatenated in field order to generate the local area time features. The local area time features from consecutive acquisition times are arranged in chronological order to generate the local area feature sequence. The same concatenation and arrangement process is performed on each adjacent partition node to generate a neighboring area feature sequence. The neighboring area weights are determined based on the partition type combination corresponding to the partition adjacent edges, where the neighboring area weight between the operation partition and the diffusion partition is 0.50. The neighboring cell weight between the zone and the access zone is 0.30, and the neighboring cell weight between the access zone and the diffusion zone is 0.20. The neighboring cell feature sequences at the same acquisition time are weighted and summed according to their neighboring cell weights to generate the neighboring cell time influence features. The neighboring cell time influence features of consecutive acquisition times are arranged in chronological order to generate the neighboring cell influence features. The local zone time features and neighboring cell time influence features at the same acquisition time are concatenated, and the concatenation result is normalized to generate the partition fusion time features of the current partition node at that acquisition time. The partition fusion time features of consecutive acquisition times are arranged in chronological order to generate the partition fusion features of the current partition node. The above processing is performed sequentially on all partition nodes to generate the partition fusion features corresponding to each partition node.

[0103] Based on the partition fusion characteristics, determine the food preparation status, cooking status, cleaning status, dining status, unattended status, and safety anomaly status of each partition node, and generate the partition scene status.

[0104] The generation of the partition scene state is as follows: Partition fusion features are processed one by one according to the partition number, and the values ​​of the corresponding kitchen safety perception data, personnel activity data, environmental perception data, and equipment operation data in the partition fusion features are uniformly converted to the range of 0 to 1. When any value of the kitchen safety perception data reaches 0.80 or above, the corresponding partition node is marked as a safety anomaly state. When the personnel activity data value is 0 and the equipment operation data value is below 0.10, the corresponding partition node is marked as an unoccupied state. When the personnel activity data value reaches 0.30 or above, the environmental perception data change value is below 0.40, and the equipment operation data value is below 0.50, the corresponding partition node is marked as a food preparation state. When the environmental perception data change value reaches 0.60 or above, and the value of the corresponding ventilation or purification operation in the equipment operation data reaches 0. When the value is above 30, the corresponding partition node is marked as cooking state; when the value of personnel activity data reaches 0.30 or above, the value of corresponding humidity change in environmental perception data reaches 0.50 or above, and the value of kitchen safety perception data is below 0.80, the corresponding partition node is marked as cleaning state; when the value of personnel activity data reaches 0.30 or above, the value of corresponding environmental perception data change is below 0.30, and the value of corresponding lighting operation in equipment operation data reaches 0.30 or above, the corresponding partition node is marked as dining state; when the same partition node meets the marking conditions for multiple states at the same time, one state mark is retained in the order of safety abnormal state, cooking state, cleaning state, dining state, food preparation state, and unattended state; the state marks of each partition node, the corresponding partition number, and the collection time are combined to generate the partition scene state.

[0105] In this embodiment, the control constraint generation module includes:

[0106] Read the partition scene status according to the partition number;

[0107] Determine the environmental improvement goals for the corresponding partition based on the status markers in the partition scenario status;

[0108] The determination of environmental improvement targets is specifically as follows: read the status markers in the partition scene state and generate environmental improvement targets according to the control requirements corresponding to the status markers; when the status marker is a safety anomaly state, generate a safety elimination target; when the status marker is a cooking state, generate a smoke suppression target and a temperature and humidity stabilization target; when the status marker is a cleaning state, generate a humidity reduction target; when the status marker is a food preparation state, generate a lighting enhancement target; when the status marker is a dining state, generate an environmental stability target; when the status marker is an unattended state, generate a low energy consumption maintenance target; and write the generated environmental improvement targets into the corresponding partition number.

[0109] Energy consumption and noise constraints for the corresponding zones are determined based on environmental improvement goals;

[0110] The determination of energy consumption constraints and noise constraints is as follows: Read the environmental improvement targets corresponding to each zone number, and match energy consumption levels and noise levels according to the environmental improvement targets; when the environmental improvement target is a safety elimination target, match the highest energy consumption level and allow for a lower noise level; when the environmental improvement target is a smoke suppression target or a temperature and humidity stabilization target, match a high energy consumption level and a wide noise level; when the environmental improvement target is a humidity reduction target or a lighting enhancement target, match a medium energy consumption level and a medium noise level; when the environmental improvement target is an environmental stabilization target, match a low energy consumption level and a low noise level; when the environmental improvement target is a low energy consumption maintenance target, match the lowest energy consumption level and a low noise level; use the energy consumption level as the energy consumption constraint for the corresponding zone, and the noise level as the noise constraint for the corresponding zone.

[0111] Determine the security priority of the corresponding partition based on the status markers and environmental improvement goals in the partition scenario state;

[0112] The safety priority is determined as follows: When the status is marked as an abnormal safety state and the environmental improvement goal is a safety elimination goal, the corresponding partition's safety priority is marked as first priority; when the status is marked as a cooking state and the environmental improvement goal includes oil fume suppression or temperature and humidity stabilization, the corresponding partition's safety priority is marked as second priority; when the status is marked as a cleaning state and the environmental improvement goal is humidity reduction, the corresponding partition's safety priority is marked as third priority; when the status is marked as a food preparation state and the environmental improvement goal is lighting enhancement, the corresponding partition's safety priority is marked as fourth priority; when the status is marked as a dining state and the environmental improvement goal is environmental stabilization, the corresponding partition's safety priority is marked as fifth priority; when the status is marked as an unoccupied state and the environmental improvement goal is low energy consumption maintenance, the corresponding partition's safety priority is marked as sixth priority; the safety priority is then written into the corresponding partition number.

[0113] The environmental improvement goals, energy consumption constraints, noise constraints, and safety priorities are combined according to the zoning number to generate a set of control constraints.

[0114] In this embodiment, the optimization decision module includes:

[0115] Read the set of control constraints and generate multiple candidate control schemes based on the set of control constraints. Each candidate control scheme includes the master control device, the coordinating device, the control amplitude, and the duration.

[0116] The generation of candidate control schemes is as follows: The environmental improvement target, energy consumption constraint, noise constraint, and safety priority are read from the control constraint set according to the zone number; kitchen environmental control equipment capable of performing corresponding control actions is matched according to the environmental improvement target; the kitchen environmental control equipment acting on the corresponding zone is marked as the main control equipment, and the kitchen environmental control equipment acting on adjacent zones and capable of cooperating with the main control equipment to achieve the same environmental improvement target is marked as the cooperating equipment; the operating range of the main control equipment and cooperating equipment is limited according to the energy consumption constraint and noise constraint; the control execution order between different zones is determined according to the safety priority; within the operating range, the main control equipment, cooperating equipment, control amplitude, and duration are combined to generate multiple candidate control schemes.

[0117] The degree to which environmental improvement goals, energy consumption constraints, noise constraints, and safety priorities are met for each candidate control scheme is used as the optimization objective. Specifically, the overlap between the operational zones of the main control equipment and the zones where the environmental improvement goals are located is statistically analyzed. The number of overlapping zones, control amplitude, and duration are weighted and accumulated to obtain the degree to which environmental improvement goals are met. The operating levels of the main control equipment and the collaborative equipment are read, and the operating level is multiplied by the duration to obtain the equipment energy consumption value. The equipment energy consumption value is compared with the allowable energy consumption level corresponding to the energy consumption constraint to obtain the degree to which the energy consumption constraint is met. The noise level corresponding to the operating level of the main control equipment and the collaborative equipment is read, and the noise level is compared with the allowable noise level corresponding to the noise constraint to obtain the degree to which the noise constraint is met. The zone number and execution order corresponding to the candidate control scheme are read, and the execution order is compared with the safety priority to obtain the degree to which the safety priority is met. The degree to which environmental improvement goals, energy consumption constraints, noise constraints, and safety priorities are met is written into the candidate control scheme as the optimization objective of the candidate control scheme.

[0118] The NSGA-III multi-objective optimization algorithm is used to perform non-dominated ranking of multiple candidate control schemes to generate the dominance level of the candidate control schemes.

[0119] The non-dominated sorting process is as follows: For each candidate control scheme, the degree of satisfaction of environmental improvement objectives, energy consumption constraints, noise constraints, and safety priority are read, and the values ​​of the four items are uniformly converted to the range of zero to one; the four satisfaction levels are reversed to generate environmental improvement objective values, energy consumption constraint objective values, noise constraint objective values, and safety priority objective values; the four objective values ​​of any two candidate control schemes are compared one by one. When one candidate control scheme has no more than the other candidate control scheme in all four objective values, and at least one objective value is less than the other candidate control scheme, the former is ranked first. A candidate control scheme is marked as dominating the next candidate control scheme; the number of times each candidate control scheme is dominated by other candidate control schemes is counted, generating a domination count; candidate control schemes with a domination count of zero are assigned to the first domination level; the domination influence of candidate control schemes in the first domination level on other candidate control schemes is deleted, and the domination count of the remaining candidate control schemes is updated; candidate control schemes with a domination count of zero after the update are assigned to the second domination level; the domination count is updated and the domination level is assigned level by level in the same way until all candidate control schemes are assigned levels, generating the domination level of the candidate control schemes;

[0120] Based on the reference point screening mechanism of the NSGA-III multi-objective optimization algorithm, reference point association and crowding screening are performed on candidate control schemes within the same dominance level to determine the target control scheme.

[0121] The determination of the target control scheme is as follows: Read the environmental improvement target value, energy consumption constraint target value, noise constraint target value, and safety priority target value of each candidate control scheme within the same dominance level, and transform these four target values ​​into the same target space; establish a set of reference points according to the number of dimensions of the four target values, distributing the reference points across the four optimization directions of environmental improvement, energy consumption, noise, and safety; calculate the distance from each candidate control scheme to each reference point, and use the reference point with the smallest distance as the associated reference point for that candidate control scheme; count the number of candidate control schemes associated with each reference point to obtain the reference point congestion count; when screening within the same dominance level, prioritize retaining candidate control schemes with higher dominance levels; when multiple candidate control schemes exist under the same partition number, select the candidate control scheme with the smaller reference point congestion count; when multiple candidate control schemes correspond to the same reference point, select the candidate control scheme with the smallest distance to that reference point; repeat the reference point association, congestion count statistics, and distance comparison until one candidate control scheme is retained for each partition number; determine the retained candidate control scheme as the target control scheme.

[0122] The main control equipment, coordinating equipment, control amplitude, and duration in the target control scheme are combined according to the partition number to generate a partition linkage control strategy.

[0123] In this embodiment, the feedback correction module includes:

[0124] Extract the master control device, cooperating devices, control amplitude and duration from the zone linkage control strategy, arrange the execution order according to safety priority, and generate equipment scheduling instructions;

[0125] Send the equipment scheduling command to the corresponding kitchen environment control equipment for execution, and obtain the controlled zoning scene status after execution;

[0126] The state of the partitioned scenario after adjustment is compared with the state of the partitioned scenario before execution to determine the state change results of each partition node.

[0127] The determination of the state change results for each partition node is as follows: Extract the state markers, environmental perception data values, personnel activity data values, and equipment operation data values ​​from both; perform a consistency comparison of the state markers before and after execution under the same partition number; when the state marker changes from a safety anomaly state, cooking state, cleaning state, or food preparation state to a dining state or unattended state, it is marked as a state improvement; when the state marker remains unchanged and the environmental perception data value decreases, it is marked as a state weakening; when the state marker remains unchanged and the environmental perception data value does not decrease, it is marked as a state not improving; when the state marker changes from a dining state or unattended state to a safety anomaly state, cooking state, cleaning state, or food preparation state, it is marked as a state worsening; write the state improvement, state weakening, state not improving, and state worsening results into the corresponding partition number to determine the state change results for each partition node.

[0128] Based on the state change results, the device action edges and partition adjacency edges in the island environment linkage diagram are corrected. Specifically, device action edge weights are written for the device action edges between the main control device, cooperating devices, and corresponding partition nodes. The value range of the device action edge weights is 0 to 1. When the state change result is state improvement, the weight of the corresponding device action edge is increased by 0.10; when the state change result is state weakening, the weight of the corresponding device action edge is increased by 0.05; when the state change result is no improvement, the weight of the corresponding device action edge is decreased by 0.10; when the state change result is state worsening, the weight of the corresponding device action edge is decreased by 0.15. Boundary constraints are applied to the corrected device action edge weights to keep them within the range of 0 to 1. The partition adjacency edges connected to the corresponding partition nodes are read. When adjacent partition nodes have the same state change result, the weight of the corresponding partition adjacency edge is increased by 0.05; when adjacent partition nodes have the opposite state change result, the weight of the corresponding partition adjacency edge is decreased by 0.05. Boundary constraints are applied to the corrected partition adjacency edge weights to complete the correction of the island environment linkage diagram.

[0129] The system combines equipment scheduling instructions, adjusted zoning scene status, and corrected island environment linkage diagram to generate intelligent control results for the kitchen island environment.

[0130] Example 1: To verify the feasibility of this invention in practice, it was applied to the environmental control scenario of an island in an open-plan kitchen in a residential building. The island is arranged in a peninsula style, with a countertop length of approximately 2.4 meters and a width of approximately 0.9 meters. One side connects to the cabinets, and the outer side faces the dining area. Before the renovation, the kitchen mainly relied on the user manually turning on the range hood, lights, and air purifier. When fumes spread to the dining area, the user usually had to notice before increasing the exhaust fan speed. When preparing food at night, even with all the kitchen lights on, the illumination on the island countertop was often not concentrated. The problems of moisture retention after cleaning and the continuous operation of equipment when no one was present were also quite obvious.

[0131] In this embodiment, the implementer deploys temperature and humidity sensors, fume sensors, air quality sensors, human presence sensors, personnel location sensors, water immersion sensors, and gas sensors above the island, outside the island, on the cabinet connection side, and on the dining room passage side. Lighting equipment, ventilation equipment, air purification equipment, temperature and humidity control equipment, and safety reminder equipment are also connected. The system runs continuously for several natural days, recording five usage scenarios daily: morning meal preparation, lunch cooking, evening dining, cleaning and tidying, and nighttime unoccupied status, with a data collection cycle of once every 5 seconds. The system preprocesses the kitchen island control-related data into a kitchen island environment dataset, dividing it into operation zones, passage zones, and diffusion zones based on island boundaries, connection patterns, and passage locations. An island environment linkage diagram is constructed, and a control constraint set is generated based on the zone scenario status. The NSGA-III multi-objective optimization algorithm is used to determine the main control device, cooperating devices, control amplitude, and duration.

[0132] In a lunchtime cooking scenario, under traditional control methods, the concentration of cooking fumes in the diffusion zone outside the island only triggers manual adjustment of the exhaust fan level after a considerable delay following the start of cooking. The peak air quality index (AQI) on the restaurant side reaches 142, and the ventilation equipment operates at high speed for 31 minutes. After implementing this invention, when both the cooking fume concentration in the operating zone and the AQI in the diffusion zone increase simultaneously, the system identifies both zones as cooking-related states, generates targets for fume suppression and temperature / humidity stability, and designates the ventilation equipment as the primary control device and the air purification equipment as a cooperating device. Continuous recording shows that the peak AQI in the diffusion zone decreases to below 96, the average high-speed operation time of the ventilation equipment is shortened to 18 minutes, and the peak noise level in the restaurant side's occupant area decreases from 56 decibels to 49 decibels.

[0133] In evening dining scenarios, traditional methods tend to perpetuate post-cooking ventilation, resulting in noticeable exhaust noise for users seated outside the island counter or near the table. This invention generates zoned scene states based on personnel activity data, environmental perception data, and equipment operation data. It identifies the operational zone and adjacent passageway zones as dining zones, sets environmental improvement targets as environmental stability targets, and re-selects equipment scheduling instructions under energy consumption and noise constraints. Records show that the system adjusted ventilation equipment from high to low intermittent operation, maintained island counter lighting on the countertop area, kept the air quality index between 78 and 86, and stabilized passageway noise below 42 decibels, an average reduction of approximately 8 decibels compared to similar periods before the modification.

[0134] In cleaning and tidying scenarios, wiping the island countertop and using the sink for short periods can cause increased humidity in the work area. Traditional methods only issue alerts when the water immersion sensor is triggered, failing to proactively regulate humidity retention. This invention records a rise in humidity in the work area from 48% to 67%, with personnel remaining in the area. Since the kitchen safety monitoring data does not meet abnormal conditions, the system identifies the corresponding area as clean, generates a humidity reduction target, and controls the temperature and humidity control devices in conjunction with the ventilation system. After adjustment, the humidity in the work area drops back to 55%, the passageway is not mistakenly triggered into cooking mode, and the safety alert devices do not issue invalid alarms.

[0135] In unmanned nighttime scenarios, the system marks each zone as unmanned based on empty personnel activity data, low equipment operating data, and stable environmental perception data, and generates a low-energy consumption target. Continuous recording shows that lighting equipment did not turn on accidentally, ventilation equipment only operated at low speeds during short-term fluctuations in air quality, and the cumulative operating time of kitchen environmental control equipment decreased by approximately 38% compared to nighttime records before the modification. Through the above applications, it is evident that this invention can complete local perception, zone identification, equipment coordination, and feedback correction around the kitchen island area. Compared to fixed control methods for the entire kitchen, it offers more accurate island scene identification, more rational equipment linkage, and better noise and energy consumption control.

[0136] Table 1. Performance Comparison of Smart Control Methods for Kitchen Island

[0137] Scene recognition accuracy (%) 84.7 91.6 Oil fume response delay (s) 74 46 Air Quality Index Peak 126 108 Average noise level during mealtimes (dB) 47.6 43.9 Tabletop illuminance stability rate (%) 86.3 92.1 Number of times safety alerts are triggered accidentally (per week) 3 2 Average daily power consumption (kWh) 2.18 1.93

[0138] As shown in Table 1, the present invention demonstrates improvements over traditional methods in several common indicators. The scene recognition accuracy increased from 84.7% to 91.6%, indicating that the present invention, through an island zoning model and island environment linkage diagram, integrates environmental perception data, personnel activity data, and equipment operation data according to operation zones, passage zones, and diffusion zones. This enables the system to more accurately distinguish between states such as food preparation, cooking, cleaning, dining, unoccupied, and safety anomalies. While traditional methods can also utilize sensor data for judgment, they often rely on single-point thresholds or the overall kitchen status, making them prone to misjudgments when personnel briefly stay, when oil fumes slightly diffuse, or when humidity changes after cleaning.

[0139] In terms of fume response and air quality control, the traditional method has a fume response delay of 74 seconds, while this invention has a delay of 46 seconds, reducing the peak air quality index from 126 to 108. This change indicates that this invention does not wait until a single fume sensor reaches a high trigger condition before controlling the equipment, but rather combines environmental changes in the operating zone and diffusion zone, as well as the adjacency of the zones, to make judgments. Therefore, it can identify the trend of fume diffusion into open spaces earlier. The decrease in the peak air quality index also indicates that ventilation equipment, air purification equipment, and other kitchen environment control equipment can work together at a more appropriate time, rather than responding independently to a single device.

[0140] In terms of average noise during dining hours, stability of tabletop illuminance, and average daily power consumption, this invention also demonstrates good overall performance. Average noise during dining hours decreased from 47.6 dB to 43.9 dB, indicating that the system can suppress unnecessary high-intensity equipment operation during dining hours; tabletop illuminance stability increased from 86.3% to 92.1%, indicating that the system can more stably adjust the island lighting according to the preparation and dining status; and average daily power consumption decreased from 2.18 kWh to 1.93 kWh, demonstrating that this invention can reduce ineffective operation while improving environmental conditions.

[0141] The number of false safety alert triggers decreased from 3 times / week to 2 times / week, indicating that the present invention does not rely solely on the instantaneous triggering of a single sensor in identifying abnormal safety conditions. Instead, it combines the status of the zoned scene and the feedback correction results for judgment, which can reduce false triggers caused by short-term humidity, personnel passing by, or changes in equipment status. Overall, the performance improvement of the present invention comes from the synergistic effect of island zone modeling, island environment linkage diagram, multi-objective optimization decision-making, and feedback correction, enabling the control of the open kitchen island area to shift from unified control of the entire kitchen to precise linkage control based on zones and scenes.

[0142] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A smart kitchen island environment control system based on the Internet of Things and sensor data, characterized in that: include: The data preprocessing module is used to collect and preprocess kitchen island control-related data to generate a kitchen island environment dataset. The island partitioning modeling module is used to divide the operation partitioning, passage partitioning, and diffusion partitioning according to the island boundary, connection form, and passage location, and to construct the island partitioning model. The linkage diagram construction module is used to construct an island environment linkage diagram based on the island zoning model, which associates the sensor coverage area and the equipment operating range. The scene state recognition module is used to generate the scene state of each zone by integrating the kitchen island environment dataset with the island environment linkage diagram. The regulation constraint generation module is used to determine environmental improvement goals, energy consumption constraints, noise constraints, and safety priorities based on the partitioned scenario state, and generate a set of regulation constraints; The optimization decision module is used to perform non-dominated sorting and reference point screening of the set of control constraints using the NSGA-III multi-objective optimization algorithm, determine the master control equipment, coordinating equipment, control amplitude and duration, and generate a zone linkage control strategy. The feedback correction module is used to generate equipment scheduling instructions based on the zone linkage control strategy, and correct the island environment linkage diagram based on the regulated zone scene status, thereby generating intelligent control results for the kitchen island environment.

2. The intelligent control system for the kitchen island environment based on the Internet of Things and sensor data according to claim 1, characterized in that, The kitchen island control-related data includes environmental perception data, personnel activity data, kitchen safety perception data, environmental control equipment operation data, and IoT device communication status data. The preprocessing includes data collection time alignment, device identification unification, data format conversion, abnormal data collection value removal, missing data collection value completion, data unit unification, and partition label writing.

3. The intelligent control system for the kitchen island environment based on the Internet of Things and sensor data according to claim 1, characterized in that, The island platform partitioning modeling module includes: Retrieve island boundary data, connection shape data, and passage location data from the kitchen island environment dataset; Map the island boundary data to the kitchen plane coordinate system to generate the island boundary coordinates; Based on the connection morphology data, the connecting side, open side, and adjacent side in the island platform boundary coordinates are marked to generate island platform boundary marking results; Based on the passage location data, the movement paths and stopping locations of people around the island platform are marked, and passage location marking results are generated; Based on the island / platform boundary marking results and access location marking results, the operation zone, access zone, and diffusion zone are divided; Generate a partition adjacency matrix based on the adjacency between the operation partition, the passage partition, and the diffusion partition; The island boundary coordinates, island boundary marking results, passage position marking results, and partition adjacency matrix are combined to construct the island partition model.

4. The intelligent control system for the kitchen island environment based on the Internet of Things and sensor data according to claim 1, characterized in that, The linkage diagram construction module includes: Based on the sensor installation location, sensing direction, and sensing distance, the sensor coverage area is matched with the boundaries of each zone to determine the sensor coverage zone; Based on the installation location, direction of action, and distance of action of the kitchen environment control equipment, the range of action of the equipment is matched with the boundaries of each zone to determine the zones in which the equipment operates; Write the operation zone, passage zone, and diffusion zone as zone nodes, write the sensors as sensing nodes, and write the kitchen environment control equipment as device nodes. Based on the sensor coverage partition, a sensing coverage edge is written between the sensing node and the partition node. The sensing coverage edge includes the main sensing coverage edge and the auxiliary sensing coverage edge. Based on the device function partition, write device function edges between the device node and the partition node. Device function edges include master device function edges and cooperating device function edges. Write partition adjacency edges between adjacent partition nodes based on the partition adjacency matrix; By combining partition nodes, sensing nodes, device nodes, sensing coverage edges, device action edges, and partition adjacency edges, an island-level environmental linkage diagram is constructed.

5. The intelligent control system for the kitchen island environment based on the Internet of Things and sensor data according to claim 1, characterized in that, The scene state recognition module includes: The system retrieves partition nodes, sensing nodes, equipment nodes, sensing coverage edges, equipment action edges, and partition adjacency edges from the island environment linkage diagram, and retrieves sensor data, equipment operation data, and personnel activity data from the kitchen island environment dataset. Based on the sensing coverage edge, sensor data is mapped to the corresponding partition node to generate partition environmental features; Based on the device action edge, the device operation data is mapped to the corresponding partition node to generate partition device characteristics; Generate zoning personnel characteristics based on personnel activity data; Based on the partition adjacency edges, the partition environment features, partition device features, and partition personnel features of adjacent partition nodes are fused to generate partition fusion features; Based on the partition fusion characteristics, determine the food preparation status, cooking status, cleaning status, dining status, unattended status, and safety anomaly status of each partition node, and generate the partition scene status.

6. The intelligent control system for the kitchen island environment based on the Internet of Things and sensor data according to claim 1, characterized in that, The regulation constraint generation module includes: Read the partition scene status according to the partition number; Determine the environmental improvement goals for the corresponding partition based on the status markers in the partition scenario status; Energy consumption and noise constraints for the corresponding zones are determined based on environmental improvement goals; Determine the security priority of the corresponding partition based on the status markers and environmental improvement goals in the partition scenario state; The environmental improvement goals, energy consumption constraints, noise constraints, and safety priorities are combined according to the zoning number to generate a set of control constraints.

7. The intelligent control system for the kitchen island environment based on the Internet of Things and sensor data according to claim 1, characterized in that, The optimization decision module includes: Read the set of control constraints and generate multiple candidate control schemes based on the set of control constraints. Each candidate control scheme includes the master control device, the coordinating device, the control amplitude, and the duration. The degree to which each candidate control scheme meets the environmental improvement target, energy consumption constraint, noise constraint, and safety priority is satisfied is used as the optimization objective. The NSGA-III multi-objective optimization algorithm is used to perform non-dominated ranking of multiple candidate control schemes to generate the dominance level of the candidate control schemes. Based on the reference point screening mechanism of the NSGA-III multi-objective optimization algorithm, reference point association and crowding screening are performed on candidate control schemes within the same dominance level to determine the target control scheme. The main control equipment, coordinating equipment, control amplitude, and duration in the target control scheme are combined according to the partition number to generate a partition linkage control strategy.

8. The intelligent control system for the kitchen island environment based on the Internet of Things and sensor data according to claim 1, characterized in that, The feedback correction module includes: Extract the master control device, cooperating devices, control amplitude and duration from the zone linkage control strategy, arrange the execution order according to safety priority, and generate equipment scheduling instructions; Send the equipment scheduling command to the corresponding kitchen environment control equipment for execution, and obtain the controlled zoning scene status after execution; The state of the partitioned scenario after adjustment is compared with the state of the partitioned scenario before execution to determine the state change results of each partition node. The device action edges and partition adjacency edges in the island environment linkage diagram are corrected based on the state change results; The system combines equipment scheduling instructions, adjusted zoning scene status, and corrected island environment linkage diagram to generate intelligent control results for the kitchen island environment.

Images