A park public area energy-saving control method based on dynamic density partitioning

Abstract

The application discloses a park public area energy-saving control method based on dynamic density partitioning, characterized in that: based on collected infrared human sensing data and / or desensitization number data, the current occupancy state of a target area is determined and classified as an unoccupied area, a low-density area or a high-density area, and a partitioning result is output; according to the partitioning result, combined with collected illumination data and temperature and humidity data, basic control instructions for lighting equipment and / or air conditioning equipment are generated; current time and date information is acquired to determine whether a preset simple mode triggering condition is met; if yes, simple mode control instructions covering all partitions are generated; if not, the basic control instructions are adopted. The application realizes differentiated energy-saving control by fusing environmental data through dynamic density partitioning, and is combined with time scene adaptation, thereby improving control granularity, tapping energy-saving potential and realizing systematic energy saving in all time periods.

Description

Technical Field

[0001] This invention belongs to the field of energy-saving technology for public areas of industrial parks, and specifically relates to an energy-saving control method for public areas of industrial parks based on dynamic density zoning. Background Technology

[0002] In the field of smart building and park energy management, the energy consumption of lighting and air conditioning systems in public areas (such as lobbies, corridors, open-plan offices, and meeting rooms) accounts for a significant portion of operating costs. To achieve energy conservation and consumption reduction, fixed timetable control based on timers is commonly used. However, its drawback is that it cannot adapt to the randomness of human activities, often resulting in energy waste when no one is present, or affecting comfort when there is unexpected human activity.

[0003] Subsequently, manned / unmanned binary sensing and control technology, represented by passive infrared (PIR) sensors, became the mainstream, realizing basic automation of lights turning on when people are present and turning off when people leave, significantly improving energy efficiency.

[0004] However, with the increasing demands for refined management and dual-carbon objectives, existing control methods based on simple binary sensing are gradually revealing their inherent technical limitations. First, their control granularity is coarse, only able to distinguish between the presence or absence of personnel. Second, environmental sensing and control strategies are often disconnected or simply combined, resulting in insufficient exploitation of energy-saving potential. Furthermore, traditional control strategies lack global adaptive capabilities based on high-level time scenarios (such as holidays or late nights), still requiring manual intervention or fixed modes, and may still exhibit standby energy consumption during prolonged periods of inactivity.

[0005] Therefore, there is a core problem that urgently needs to be solved: how to overcome energy waste in various scenarios and achieve coordinated optimization control of environmental parameters and personnel status, so as to further explore energy-saving potential while ensuring basic comfort. Summary of the Invention

[0006] This invention provides an energy-saving control method for public areas in a park based on dynamic density zoning. By integrating infrared human sensing and desensitized visual data, it dynamically divides unmanned, low-density, and high-density areas, and generates differentiated lighting and air conditioning control strategies based on the zoning results and environmental parameters. At the same time, it superimposes time scene modes for global optimization to solve the energy waste problem caused by existing scenarios and overcome the technical defects of insufficient coordination between environmental control and personnel status and time cycle.

[0007] The technical solution adopted in this invention is as follows: An energy-saving control method for public areas of a park based on dynamic density zoning, comprising: Based on the collected infrared human sensing data and / or desensitized human data, determine the current occupancy status of the target area and classify it into uninhabited area, low-density area or high-density area, and output the zoning results. Based on the zoning results and combined with the collected illuminance and temperature and humidity data, basic control commands for lighting equipment and / or air conditioning equipment are generated. Obtain the current time and date information, and determine whether the preset simple mode triggering conditions are met; if they are met, generate simple mode control instructions covering all partitions; if they are not met, use the basic control instructions.

[0008] The energy-saving control method for public areas of a park based on dynamic density zoning adopted in this invention also has the following additional technical features: The collection of anonymized data includes: The acquired image frames are processed in real time within the memory of the edge vision device; The image frame is analyzed using a pre-set people recognition model, and the output count value without personal identification information is used as de-identified people data. After completing the population analysis, the original data of the image frame is discarded.

[0009] Determining the current occupancy status of the target area specifically includes: When the infrared human detection data indicates no human activity and the desensitized human data is zero, it is determined to be an uninhabited area; When the infrared human sensing data indicates human activity but the trigger frequency is lower than the first threshold, or when the desensitized human data is not zero and is lower than the second threshold, it is determined to be a low-density area. When the infrared human sensing data indicates human activity but the trigger frequency is greater than or equal to the first threshold, or when the desensitized human data is greater than or equal to the second threshold, it is determined to be a high-density area.

[0010] Generate basic control commands, specifically including: When the partitioning result is an uninhabited area, a lighting off command and an air conditioning standby command are generated; When the zoning result is a low-density area, an instruction is generated to adjust the lighting brightness to the first preset ratio, and an instruction is generated to increase the air conditioning set temperature by the first preset value based on the standard set value. When the zoning result is a high-density area, at least a lighting brightness adjustment command is generated based on the illuminance data, and an air conditioning operation command is generated based on the temperature and humidity data.

[0011] Based on the illuminance data, a lighting brightness adjustment command is generated, specifically as follows: When the illuminance data is below the third threshold, the lighting brightness is controlled to 100%; otherwise, the lighting brightness is reduced according to the illuminance data.

[0012] Preset simple mode trigger conditions, such as holidays or nighttime hours; Simple mode control commands include reducing the lighting brightness of all areas to the minimum safe brightness, and controlling the air conditioning in all areas to enter off or ventilation standby mode.

[0013] The method is executed cyclically at a fixed time period to achieve real-time dynamic adjustment.

[0014] A second aspect of the present invention provides an energy-saving control system for public areas of a park based on dynamic density zoning, for implementing the method described above, the system comprising: The sensing module is configured to collect infrared human sensing data, illuminance data, temperature and humidity data of the target area, as well as desensitized human data through the edge vision analysis unit. A data processing and control decision module, communicatively connected to the sensing module, is configured to receive data from the sensing module and execute the dynamic density zoning and strategy generation logic as described in any one of claims 2 to 7; output control commands; and an execution module, communicatively connected to the data processing and control decision module, is configured to receive the control commands and drive the corresponding lighting and air conditioning equipment to operate.

[0015] The edge visual analysis unit includes: Image acquisition unit; The edge computing chip has a built-in people recognition model, which is used to process the image frames captured by the image acquisition unit and output de-identified people data; The edge computing chip is configured not to store the original image data after processing.

[0016] The system also includes an interface module for communicating with the management platform, used to receive configuration information of the simple mode triggering conditions and upload energy consumption and operating status data.

[0017] Due to the adoption of the above technical solution, the beneficial effects achieved by this invention are as follows: 1. In this invention, firstly, by dynamically zoning (distinguishing between unmanned areas, low-density areas, and high-density areas), the limitations of traditional manned / unmanned binary control are overcome. It can accurately identify low-density states with sparse personnel, thus providing a decision-making basis for implementing differentiated energy-saving strategies. This fundamentally avoids energy waste caused by operating equipment at full power in scenarios with only a few personnel present, achieving a refined leap in the granularity of energy-saving control.

[0018] Secondly, when generating control commands, the zoning results are combined with real-time illuminance, temperature, and humidity data, enabling the lighting and air conditioning control strategies to not only respond to population density but also dynamically adapt to environmental conditions. This allows the system to fully utilize natural light and avoid excessive cooling or heating while ensuring basic comfort, further unlocking energy-saving potential.

[0019] By introducing a simple pattern judgment and overriding mechanism based on time and date, the system gains global scene adaptability. During long periods of inactivity, such as nighttime and holidays, this method can automatically switch to the lowest energy consumption global management mode, effectively eliminating the standby power consumption or management blind spots that may occur in traditional systems due to zone control during non-working hours, and achieving systemic energy saving across all time periods and all areas. Attached Figure Description

[0020] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this invention, illustrate exemplary embodiments of the invention and are used to explain the invention, but do not constitute an undue limitation of the invention. In the drawings: Figure 1 This is a flowchart illustrating the energy-saving control method for public areas of a park based on dynamic density zoning, according to one embodiment of the present invention. Detailed Implementation

[0021] To more clearly illustrate the overall concept of the present invention, a detailed description will be provided below with reference to the accompanying drawings and examples.

[0022] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0023] like Figure 1 As shown, an energy-saving control method for public areas in a park based on dynamic density zoning includes: S100: Based on the collected infrared human sensing data and / or desensitized human data, determine the current occupancy status of the target area and classify it into uninhabited area, low-density area or high-density area, and output the zoning results.

[0024] The main purpose of this step is to refine the identification and classification of the occupancy status of the target area, breaking through the traditional simple binary judgment of whether it is occupied or unoccupied.

[0025] In practice, two types of sensing data are collected synchronously or periodically: one is infrared human sensing data generated by passive infrared sensors deployed in the area, which can reflect the presence and frequency of human heat source movement; the other is, optionally, desensitized human head data provided by compliant edge vision devices, which is obtained by localized analysis of real-time video streams, and only outputs the head count results to ensure that the original image information is not retained.

[0026] By fusing and analyzing these two types of data (for example, relying on the data when only infrared sensing is enabled; and performing data complementation and verification when both methods are enabled simultaneously), the current occupancy status of the target area is dynamically determined and classified into one of three levels: uninhabited area, low-density area, or high-density area, and the partitioning result is output.

[0027] This step provides a precise basis for decision-making in implementing differentiated energy-saving strategies that match population density.

[0028] S200: Based on the partitioning results and combined with the collected illuminance data and temperature and humidity data, generate basic control commands for lighting equipment and / or air conditioning equipment.

[0029] The main purpose of this step is to generate corresponding, optimized device control commands based on the accurate partitioning results and in combination with real-time environmental parameters.

[0030] After obtaining the zoning results for uninhabited areas, low-density areas, or high-density areas, the current environmental illuminance and temperature / humidity data will be further read. The generation of control strategies follows a zoning-oriented principle: for uninhabited areas, the instructions aim to achieve the lowest energy consumption, such as controlling lighting to be completely turned off, and air conditioning to enter low-power standby or turn off state; For low-density areas, the instructions aim to balance comfort and energy conservation, such as controlling lighting to a suitable lower brightness level while adjusting the air conditioning set temperature accordingly to save energy. For high-density areas, the instructions prioritize ensuring a comfortable environment and incorporate environmental parameters for optimization. For example, lighting brightness needs to be dynamically compensated and adjusted in conjunction with natural light intensity, and air conditioning operation is precisely controlled based on the current temperature and humidity.

[0031] By combining the dimension of personnel density zoning with the environmental dimensions of light, temperature and humidity to generate control commands, it is possible to move from simply responding to the presence of personnel to coordinating and optimizing the human-environment relationship. Under the premise of meeting basic environmental needs, it is possible to deeply explore the energy-saving potential of various density scenarios.

[0032] S300: Obtain the current time and date information, and determine whether the preset simple mode triggering conditions are met; if met, generate a simple mode control instruction covering all partitions; if not met, use the basic control instruction.

[0033] The main purpose of this step is to introduce a higher-level time-dimensional strategy to give the system global scene adaptive capabilities in order to achieve systemic energy saving throughout the entire time period.

[0034] Upon receiving basic control commands, the system does not issue them immediately. Instead, it first obtains the current system time and calendar information and determines whether they meet the preset simple mode trigger conditions. This mode is designed to cover global scenarios where long periods of inactivity or minimal human activity are expected, such as nighttime or holidays.

[0035] If the triggering conditions are met, the system will ignore the specific zoning results for each area and uniformly generate and issue a set of simple mode control commands covering all zones (for example, uniformly adjust the lighting of the entire area to the lowest safe brightness, and uniformly switch the air conditioning to off or ventilation standby mode). If the triggering conditions are not met, the system will directly use basic control commands. Finally, the selected control commands are issued to the corresponding lighting and air conditioning actuators.

[0036] This step establishes a two-tier decision-making system based on basic zoning strategies and global time strategies. It effectively solves the problem of fragmented standby power consumption that may still exist during non-working hours in zoning control based on real-time perception. Through a mandatory global energy-saving coverage, it ensures that the overall energy-saving effect can be maximized during the inactive period of the park, and realizes closed-loop optimization of energy management.

[0037] As a preferred embodiment of the present invention, the collection of anonymized data includes: The acquired image frames are processed in real time within the memory of the edge vision device; The image frame is analyzed using a pre-set people recognition model, and the output count value without personal identification information is used as de-identified people data. After completing the population analysis, the original data of the image frame is discarded.

[0038] This implementation method collects anonymized human data to achieve high-precision dynamic density zoning while strictly adhering to privacy protection. The core objective is to improve the accuracy of human density assessment using visual information without infringing on individual privacy, thereby supporting more refined energy-saving control.

[0039] First, data acquisition and edge processing: Image acquisition is handled by edge vision devices (such as smart cameras or edge computing boxes with computing capabilities) deployed in the target area. Unlike traditional cloud-based video analytics, raw image data is strictly confined to the device's local memory and is not transmitted to external servers over the network, physically cutting off the path for privacy data leakage.

[0040] Secondly, real-time analysis and information anonymization: In the device's memory, a pre-built, lightweight people recognition model (e.g., a deep learning-based neural network model focused on detecting human contours or head and shoulder features) analyzes the real-time captured image frames. This model is specially trained and optimized, and its output consists only of pure count values ​​without any personally identifiable features (such as facial features or clothing details). This process completes the anonymization process.

[0041] Finally, data destruction and compliance assurance: Once the analysis task is completed, critical data cleanup operations are immediately performed, discarding or overwriting the original image frame memory data used for analysis. This means that only the anonymized headcount figures, presented as analysis results, are retained at the edge. This not only improves the level of anonymization but also reduces storage size, making it suitable for timely edge processing.

[0042] This implementation overcomes the potential for missed detections (such as stationary people) or misjudgments that may occur when relying solely on infrared human sensors. By using visual confirmation, it significantly improves the reliability of determining unmanned, low-density, and high-density states, providing a more reliable input for subsequent differentiated energy-saving strategies.

[0043] As a preferred embodiment of the present invention, determining the current occupancy status of the target area specifically includes: When the infrared human detection data indicates no human activity and the desensitized human data is zero, it is determined to be an uninhabited area; When the infrared human sensing data indicates human activity but the trigger frequency is lower than the first threshold, or when the desensitized human data is not zero and is lower than the second threshold, it is determined to be a low-density area. When the infrared human sensing data indicates human activity but the trigger frequency is greater than or equal to the first threshold, or when the desensitized human data is greater than or equal to the second threshold, it is determined to be a high-density area.

[0044] In this implementation method, determining the current occupancy status of the target area is crucial for constructing a dynamic density zoning model and achieving refined sensing. Its main objective is to accurately map abstract sensing data into three physical states with clear energy-saving guidance significance: uninhabited areas, low-density areas, and high-density areas, by formulating clear and executable multi-source data fusion judgment rules. This provides precise and reliable input for subsequent differentiated control.

[0045] Specifically, this determination process is implemented through the following strategies and steps. First, a determination framework based on a combination of thresholds and logic is established. The inputs to this framework are real-time acquired infrared human sensing data and optional desensitized human data. The infrared human sensing data not only provides binary information indicating presence or absence, but its trigger frequency within a certain time window is also quantified as a continuous indicator characterizing the intensity of human activity. The desensitized human data directly provides the absolute number of people confirmed visually.

[0046] Secondly, a tiered judgment logic is implemented. First, a no-man's-land is determined: this determination is made only when infrared human detection data consistently indicates the absence of any human activity and the visual perception channel (if enabled) confirms zero people. This effectively prevents false no-man's-land judgments due to missed detections by a single sensor (e.g., infrared sensors are insensitive to stationary people), avoids accidentally shutting down equipment while people are still present, and ensures basic comfort and safety.

[0047] Second, it is determined to be a low-density area: it covers two typical low-density scenarios. One is where people are present but their activities are sparse, which is manifested by the infrared human detection trigger frequency being lower than the preset first threshold; the other is where people are visually perceived, but the number of people is lower than the preset second threshold.

[0048] These two thresholds can be set empirically or calibrated adaptively based on the size and function of the specific area. This strategy can accurately detect scenes with sparse population and low activity.

[0049] Third, it is classified as a high-density area: when infrared human detection data shows frequent human activity (trigger frequency greater than or equal to the first threshold), or when the number of people perceived visually reaches or exceeds a certain scale (greater than or equal to the second threshold), it is classified as this state. This indicates an active area where environmental comfort needs to be prioritized.

[0050] By introducing frequency and number of people thresholds, the perception of simple "occupant" status is further refined, thus subdividing the "occupant" status, which is a prerequisite for implementing differentiated energy-saving strategies. Utilizing the complementarity of infrared human sensing and visual data (e.g., infrared is sensitive to motion, while vision is effective for static presence), cross-validation through logical combination significantly improves the overall accuracy of status determination and reduces the false positive rate.

[0051] Specifically, generating basic control commands includes: When the partitioning result is an uninhabited area, a lighting off command and an air conditioning standby command are generated; When the zoning result is a low-density area, an instruction is generated to adjust the lighting brightness to the first preset ratio, and an instruction is generated to increase the air conditioning set temperature by the first preset value based on the standard set value. When the zoning result is a high-density area, at least a lighting brightness adjustment command is generated based on the illuminance data, and an air conditioning operation command is generated based on the temperature and humidity data.

[0052] For three different physical scenarios—uninhabited areas, low-density areas, and high-density areas—we formulate and implement precisely matched, quantified equipment control strategies to achieve on-demand energy allocation and reduce waste while ensuring basic environmental comfort.

[0053] Specifically, regarding instruction generation for unoccupied areas: when the zoning result confirms that the area is unoccupied, the core of the control strategy is to eliminate all unnecessary standby power consumption. Explicit instructions for turning off lighting and activating air conditioning standby will be generated. The standby instructions for air conditioning can be specified based on the device type, such as turning off the compressor, entering low-speed ventilation, or keeping the water valve closed—all low-power states—eliminating energy consumption in the area during periods of no demand.

[0054] Command generation for low-density areas: In this scenario, there is human activity, but the intensity of demand is low. Two sets of key commands will be generated: First, a lighting control command, which adjusts the brightness of the lights to a first preset ratio (e.g., 50%), significantly reducing lighting energy consumption while providing basic visibility; second, an air conditioning control command, which, based on the standard set temperature (e.g., 26℃) that meets the basic thermal comfort of personnel, raises the set temperature by a first preset value (e.g., 2℃, i.e., set to 28℃).

[0055] Taking advantage of the human body's wider temperature tolerance range under low-activity conditions and the visual adaptability in low-light environments, significant energy savings in lighting and air conditioning systems are achieved with almost no impact on the user's subjective experience.

[0056] Command generation for high-density areas: This scenario prioritizes comfort and functional efficiency while pursuing energy efficiency optimization within a comfortable range. At least two aspects of fine-tuning should be implemented: First, in lighting control, brightness adjustment commands should be generated based on real-time collected illuminance data. For example, an illuminance threshold can be set, automatically reducing artificial lighting brightness as a supplement when natural light is sufficient, and increasing brightness when natural light is insufficient, avoiding leaving all lights on even in bright sunlight.

[0057] Secondly, in terms of air conditioning control, operating instructions are generated based on real-time temperature and humidity data, enabling the air conditioning system to respond more accurately and quickly to environmental changes, avoiding overcooling or overheating, and improving operating efficiency while stabilizing environmental quality.

[0058] By incorporating environmental parameters such as light intensity, temperature, and humidity for closed-loop optimization, the system can avoid ineffective energy overflow through dynamic fine-tuning even under full-load demand scenarios. This demonstrates energy efficiency management across all operating conditions and scenarios, and improves the level of intelligent control in high-density areas.

[0059] Specifically, based on the illuminance data, a lighting brightness adjustment command is generated, which includes: When the illuminance data is below the third threshold, the lighting brightness is controlled to 100%; otherwise, the lighting brightness is reduced according to the illuminance data.

[0060] This embodiment achieves intelligent and refined management of the lighting system, maximizing the use of natural light and reducing energy consumption of artificial lighting while ensuring visual comfort. Its main purpose is to establish an automated, responsive lighting control mechanism that allows the intensity of artificial lighting to be dynamically compensated according to real-time natural lighting conditions, thus eliminating energy waste caused by excessive lighting.

[0061] A third threshold is preset, which represents the minimum ambient illuminance value required to ensure the basic visual operation requirements of the area (for example, the general desktop illuminance requirement in an office area is 300 lux, and this value can be used as a reference setting point).

[0062] When the real-time collected illuminance data is lower than the third threshold, it indicates that natural light is insufficient and cannot meet the basic lighting needs of the area. At this time, an instruction is generated to control the artificial lighting in the area to operate at 100% of its rated brightness to completely compensate for the lack of natural light and ensure that the ambient brightness meets the standards.

[0063] Furthermore, it should be noted that when the real-time collected illuminance data is higher than or equal to the third threshold, it indicates that natural light is sufficient to meet or even exceed basic lighting needs. At this point, an instruction is generated to dynamically reduce the brightness of artificial lighting based on the illuminance data.

[0064] The adjustment strategy can be: graded dimming, setting multiple higher illuminance thresholds, and adjusting the artificial lighting brightness to a lower percentage (such as 70%, 50%, 30% until it is turned off) when the natural light is more abundant.

[0065] Continuous dimming establishes a compensation algorithm model (e.g., a linear or nonlinear negative correlation) between natural light illuminance and the required artificial lighting brightness, achieving stepless smooth adjustment. For example, if the system aims to maintain a constant total desktop illuminance at a third threshold level, when natural light provides 200 lux, artificial lighting only needs to supplement 100 lux. The corresponding dimming percentage is calculated, and control commands are generated.

[0066] This method can automatically balance indoor illuminance and reduce indoor brightness fluctuations caused by changes in natural light (such as cloud cover and changes in light intensity between day and night), providing users with a more uniform and stable visual environment and improving the comfort of the lighting environment.

[0067] In a preferred embodiment of the present invention, the method is executed cyclically at a fixed time period to achieve real-time dynamic adjustment.

[0068] Specifically, after initialization or startup, a control cycle loop defined by a fixed duration is entered. Within each complete control cycle (e.g., it can be set to 10 seconds, 30 seconds, or 60 seconds), the entire process from data acquisition, status determination, policy generation to instruction issuance is executed sequentially.

[0069] At the start of each cycle, the system synchronously or sequentially collects all necessary sensor data (infrared human detection, illuminance, temperature and humidity) and processes visual data; then, based on this latest data, it immediately performs dynamic density zoning determination; next, it generates or updates control commands based on the zoning results and environmental data; finally, it sends the commands to the execution devices.

[0070] After the current cycle ends, the system starts the next identical cycle immediately after a very short interval or without an interval, and so on. The fixed time period can be configured according to the response speed requirements and system processing capacity of the actual application scenario to balance real-time performance and system overhead.

[0071] The second invention provides an energy-saving control system for public areas of a park based on dynamic density zoning, used to implement the method described above. The system includes: The sensing module is configured to collect infrared human sensing data, illuminance data, temperature and humidity data of the target area, as well as desensitized human data through the edge vision analysis unit. A data processing and control decision module, communicatively connected to the sensing module, is configured to receive data from the sensing module and execute the dynamic density zoning and strategy generation logic as described in any one of claims 2 to 7; output control commands; and an execution module, communicatively connected to the data processing and control decision module, is configured to receive the control commands and drive the corresponding lighting and air conditioning equipment to operate.

[0072] The edge visual analysis unit includes: Image acquisition unit; The edge computing chip has a built-in people recognition model, which is used to process the image frames captured by the image acquisition unit and output de-identified people data; The edge computing chip is configured not to store the original image data after processing.

[0073] The system also includes an interface module for communicating with the management platform, used to receive configuration information of the simple mode triggering conditions and upload energy consumption and operating status data.

[0074] Therefore, it can achieve any effect in the energy-saving control method of public areas in the park based on dynamic density zoning, which will not be elaborated here.

[0075] For any parts not mentioned in this invention, existing technologies can be used or referenced.

[0076] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0077] The above description is merely an embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of the present invention should be included within the scope of the claims of the present invention.

Claims

1. A method for energy-saving control of public areas in a park based on dynamic density zoning, characterized in that, include: Based on the collected infrared human sensing data and / or desensitized human data, determine the current occupancy status of the target area and classify it into uninhabited area, low-density area or high-density area, and output the zoning results. Based on the partitioning results and combined with the collected illuminance and temperature and humidity data, basic control commands for lighting equipment and / or air conditioning equipment are generated. Obtain the current time and date information, and determine whether the preset simple mode trigger conditions are met; If the conditions are met, a simple mode control instruction covering all partitions is generated; otherwise, the basic control instruction is used.

2. The method according to claim 1, characterized in that, The collection of anonymized data includes: The acquired image frames are processed in real time within the memory of the edge vision device; The image frame is analyzed using a pre-set people recognition model, and the output count value without personal identification information is used as de-identified people data. After completing the population analysis, the original data of the image frame is discarded.

3. The method according to claim 1, characterized in that, Determining the current occupancy status of the target area specifically includes: When the infrared human detection data indicates no human activity and the desensitized human data is zero, it is determined to be an uninhabited area; When the infrared human sensing data indicates human activity but the trigger frequency is lower than the first threshold, or when the desensitized human data is not zero and is lower than the second threshold, it is determined to be a low-density area. When the infrared human sensing data indicates human activity but the trigger frequency is greater than or equal to the first threshold, or when the desensitized human data is greater than or equal to the second threshold, it is determined to be a high-density area.

4. The method according to claim 3, characterized in that, Generate basic control commands, specifically including: When the partitioning result is an uninhabited area, a lighting off command and an air conditioning standby command are generated; When the zoning result is a low-density area, an instruction is generated to adjust the lighting brightness to the first preset ratio, and an instruction is generated to increase the air conditioning set temperature by the first preset value based on the standard set value. When the zoning result is a high-density area, at least a lighting brightness adjustment command is generated based on the illuminance data, and an air conditioning operation command is generated based on the temperature and humidity data.

5. The method according to claim 4, characterized in that, Based on the illuminance data, a lighting brightness adjustment command is generated, specifically as follows: When the illuminance data is below the third threshold, the lighting brightness is controlled to 100%; otherwise, the lighting brightness is reduced according to the illuminance data.

6. The method according to claim 1, characterized in that, Preset simple mode trigger conditions, such as holidays or nighttime hours; Simple mode control commands include reducing the lighting brightness of all areas to the minimum safe brightness, and controlling the air conditioning in all areas to enter off or ventilation standby mode.

7. The method according to any one of claims 1 to 6, characterized in that, The method is executed cyclically at a fixed time period to achieve real-time dynamic adjustment.

8. An energy-saving control system for public areas of a park based on dynamic density zoning, characterized in that, The system for implementing the method as described in any one of claims 1 to 7 comprises: The sensing module is configured to collect infrared human sensing data, illuminance data, temperature and humidity data of the target area, as well as desensitized human data through the edge vision analysis unit. A data processing and control decision module, communicatively connected to the sensing module, is configured to receive data from the sensing module and execute the dynamic density zoning and strategy generation logic as described in any one of claims 2 to 7; output control commands; and an execution module, communicatively connected to the data processing and control decision module, is configured to receive the control commands and drive the corresponding lighting and air conditioning equipment to operate.

9. The system according to claim 8, characterized in that, The edge visual analysis unit includes: Image acquisition unit; The edge computing chip has a built-in people recognition model, which is used to process the image frames captured by the image acquisition unit and output de-identified people data; The edge computing chip is configured not to store the original image data after processing.

10. The system according to claim 8 or 9, characterized in that, The system also includes an interface module for communicating with the management platform, used to receive configuration information of the simple mode triggering conditions and upload energy consumption and operating status data.

Images