Hotel energy optimization management method and system based on group intelligence

Abstract

The invention belongs to the technical field of maintenance operation management, and particularly relates to a hotel energy optimization management method and system based on swarm intelligence, and the method comprises the steps: obtaining an echo signal through a millimeter wave radar in a guest room, carrying out the preprocessing of the echo signal, locking a target signal source, and extracting a time domain displacement signal; constructing a frequency domain energy concentration degree index and a time domain rhythm stability index based on the time domain displacement signal; performing weighted linear fusion on the frequency domain energy aggregation degree index and the time domain rhythm stability index, and introducing an environment bottom noise offset parameter for correction to obtain living body confidence comprehensively reflecting the strength of the target biological attribute; and comparing the living body confidence coefficient with a preset judgment threshold value, if the living body confidence coefficient is greater than the judgment threshold value, judging that a person is in the guest room and maintaining the current energy mode, and if the living body confidence coefficient is smaller than the judgment threshold value, judging that no person is in the guest room and executing an energy-saving program. The hotel energy management efficiency is improved.

Description

Technical Field

[0001] This invention relates to the field of maintenance, repair, operation, and management technology. More specifically, this invention relates to a hotel energy optimization management method and system based on swarm intelligence. Background Technology

[0002] With the deepening of dual-carbon goals, the hotel industry, as a key area of ​​building energy consumption, faces increasingly urgent needs for energy conservation and emission reduction. In guest room energy management, accurately identifying whether someone is in the room is a prerequisite for achieving automated energy-saving control of equipment such as air conditioning and lighting. Currently, mainstream solutions generally use card-operated power switches or infrared pyroelectric sensors for judgment: the former relies on the guest actively removing the card, and cannot trigger a power cut-off in scenarios where the person has left but the card has not been removed, resulting in continuous energy consumption in empty rooms; the latter only responds to moving heat sources and is insensitive to guests in a sleeping or stationary state, easily misjudging that no one is there and prematurely turning off the air conditioning, seriously affecting guest comfort and satisfaction.

[0003] To overcome the aforementioned limitations, the industry has gradually introduced millimeter-wave radar liveness detection technology. This solution is based on millimeter-wave radar and captures the micro-Doppler effect caused by human breathing and heartbeat to achieve accurate perception of people in a stationary state. When the system detects a signal that matches human physiological characteristics, it determines that there is someone in the room and maintains normal operation of the equipment. This method performs well in the laboratory or under ideal conditions with low interference and can effectively solve the problem that traditional technologies cannot see stationary human bodies.

[0004] However, this technology faces serious challenges in real hotel room environments. Guest rooms are not static, enclosed spaces; they often contain various non-stationary environmental interference sources, such as air conditioning vents blowing curtains, swaying plant leaves, slight swaying of paintings, and mechanical movements like the rotation of electric fans. These minute vibrations from non-human targets can also generate low-frequency Doppler signals in radar echoes, whose frequency components highly overlap with human breathing frequencies. Existing radar signal processing algorithms often employ simple spectral energy threshold mechanisms, meaning that if the signal energy exceeds a preset threshold within a specific frequency band, it is considered to be occupied. Such methods lack the ability to identify the microstructure of signals and cannot distinguish between physiologically rhythmic human micro-movements and random or quasi-periodic environmental mechanical micro-movements.

[0005] This leads to the problem of micro-motion feature confusion. When the air conditioner blows the curtains and produces a continuous low-frequency sway, the system is very likely to misjudge it as a stationary human body. As a result, the air conditioner and lighting are kept on even when the room is actually empty, forming a false target and causing continuous hidden energy waste. This problem not only weakens the energy-saving benefits of millimeter-wave radar technology, but may also lead hotels into a dilemma of high investment and poor results in intelligentization. Therefore, there is an urgent need for a highly robust sensing method that can effectively decouple human physiological signals from environmental interference, so as to truly realize refined guest room energy management that provides energy when people are present and saves energy when people leave. Summary of the Invention

[0006] To address the technical problems of misjudgment of living individuals and hidden energy consumption caused by the overlap of environmental interference and human respiratory signal spectrum in hotel rooms, the present invention provides solutions in the following aspects.

[0007] In a first aspect, the present invention provides a hotel energy optimization management system based on swarm intelligence. This includes: acquiring echo signals using millimeter-wave radar deployed in guest rooms; preprocessing the echo signals to locate the target signal source and extract the time-domain displacement signal; obtaining a frequency-domain energy concentration index based on the time-domain displacement signal, where the frequency-domain energy concentration index is the ratio of the total energy within a specific bandwidth near the dominant frequency point in the power spectrum of the time-domain displacement signal to the total energy within the filter passband, used to evaluate the purity of the signal frequency components; and obtaining a time-domain rhythm stability index based on the time-domain displacement signal, where the time-domain rhythm stability index is based on multiple peaks in the autocorrelation function of the time-domain displacement signal. The ratio of the correlation value to its zero-delay autocorrelation value is used to evaluate the signal's ability to maintain rhythm in the time dimension. Based on a weighted linear model, the frequency domain energy concentration index and the time domain rhythm stability index are weighted and linearly fused, and an environmental noise bias parameter is introduced for correction to obtain a liveness confidence score that comprehensively reflects the strength of the target's biological attributes. The liveness confidence score is compared with a preset judgment threshold. If the liveness confidence score is greater than the judgment threshold, it is determined that there is someone in the room and the current energy mode is maintained. If the liveness confidence score is less than the judgment threshold, it is determined that there is no one in the room and an energy-saving program is executed.

[0008] This invention extracts deep features from two dimensions of millimeter-wave radar echo signals: frequency domain energy concentration index and time domain rhythm stability index. This fundamentally decouples human physiological micro-movements from environmental mechanical interference, effectively solving the problem of micro-movement feature confusion caused by the swaying of non-stationary targets such as curtains and green plants in existing technologies. Instead of solely relying on echo energy magnitude, this method utilizes a weighted linear model and introduces an environmental noise bias parameter to obtain liveness confidence, significantly reducing the false alarm rate caused by environmental noise. This not only significantly improves the precision of hotel energy management but also ensures highly robust perception of energy supply and energy conservation as soon as guests are present, avoiding the impact of misjudgments on guest comfort.

[0009] Preferably, the millimeter-wave radar is deployed in the center of the guest room ceiling, ensuring no blind spots in the guest room through multi-angle radiation coverage.

[0010] Preferably, the preprocessing of the echo signal includes mixing and sampling the echo signal to obtain an intermediate frequency signal, then performing a fast Fourier transform along the fast time axis to generate a range Doppler heatmap; after filtering out static clutter based on a background cancellation algorithm, the range cell with the strongest energy is locked as the target signal source to be analyzed, the phase signal of the range cell is extracted, and the phase signal of the range cell is unwound and bandpass filtered to obtain a time-domain displacement signal reflecting the micro-motion characteristics of the target.

[0011] This invention extracts a pure time-domain displacement signal from the original echo mixed with static clutter and thermal noise through fast Fourier transform processing, unwinding, and bandpass filtering. This provides a structured underlying support for subsequent microscopic feature extraction and prevents misjudgment caused by signal distortion.

[0012] Preferably, the expression for the frequency domain energy concentration index is: In the formula, It is an index of energy concentration in the frequency domain; The power spectral density is obtained through high-resolution spectral analysis; This is the dominant frequency point with the highest energy in the power spectrum; Half of the main frequency bandwidth window; and These are the low-end cutoff frequency and the high-end cutoff frequency of the bandpass filter, respectively. It is the numerical stability constant; This represents the total energy within a specific bandwidth near the dominant frequency. This represents the total energy within the filter's passband.

[0013] This invention utilizes a normalized evaluation mechanism to assess the frequency domain energy concentration index, effectively identifying single-frequency signals that conform to human physiological characteristics, and classifying multi-frequency spurious or broadband noise as environmental interference, thus achieving preliminary physical property decoupling.

[0014] Preferably, the expression for the time-domain rhythm stability index is: In the formula, As an index of temporal rhythm stability; The autocorrelation function of the extracted time-domain displacement signal; This is the zero-delay autocorrelation value of the signal; The first autocorrelation function The correlation value corresponding to each peak; For the autocorrelation function, the th The delay time corresponding to each peak; The number of selected autocorrelation peaks; It is a natural constant; This is the preset attenuation penalty coefficient.

[0015] This invention measures rhythm persistence by introducing an attenuation penalty coefficient and overcomes the interference of quasi-periodic mechanical oscillations on the system by utilizing the rhythm maintenance ability of human respiratory height, thus providing a reliable basis for determining the stability of time-domain rhythms.

[0016] Preferably, the expression for the liveness confidence level is: In the formula, For liveness confidence; It is an index of energy concentration in the frequency domain; The signal-to-noise ratio of the current distance cell; This is the environmental noise floor bias parameter; Preset weighting coefficients; This is the kurtosis adjustment factor for the Sigmoid function; It is an index for the stability of time-domain rhythms.

[0017] Preferably, the preset weighting coefficient .

[0018] Preferably, the energy-saving procedure includes gradually reducing the air conditioning load or cutting off unnecessary power.

[0019] Preferably, the method further includes uploading the determination result as a perception feedback to a cloud-based swarm intelligence system, whereby the system dynamically adjusts the environmental noise bias parameter and determination threshold in the room's algorithm based on the room's historical false alarm rate.

[0020] This invention drives a cloud-based swarm intelligence system through perception feedback, dynamically optimizing environmental noise bias parameters and judgment thresholds using historical false alarm data. This enables the algorithm to continuously evolve with changes in room type and environment, improving the system's universality and long-term operational stability.

[0021] In a second aspect, the present invention provides a hotel energy optimization management system based on swarm intelligence, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned hotel energy optimization management method based on swarm intelligence is implemented.

[0022] By adopting the above technical solution, the above-mentioned hotel energy optimization management method based on swarm intelligence is generated into a computer program and stored in a memory so that it can be loaded and executed by a processor. In this way, a terminal device can be made based on the memory and the processor for convenient use.

[0023] The beneficial effects of this invention are as follows: By extracting the deep features of two dimensions—frequency domain energy concentration index and time domain rhythm stability index—from millimeter-wave radar echo signals, this invention fundamentally decouples human physiological micro-movements from environmental mechanical interference, effectively solving the problem of micro-movement feature confusion caused by the swaying of non-stationary targets such as curtains and green plants in existing technologies. This method no longer relies solely on the echo energy magnitude, but instead uses a weighted linear model and introduces environmental noise bias parameter correction to obtain the liveness confidence, greatly reducing the false alarm rate caused by environmental noise. While significantly improving the level of precision in hotel energy management, it ensures a highly robust perception of energy supply when people are present and energy conservation when people leave, avoiding the impact on guest comfort caused by misjudgment. Attached Figure Description

[0024] Figure 1 This is a flowchart illustrating the hotel energy optimization management method based on swarm intelligence in this invention; Figure 2 This is a frequency domain energy concentration index analysis chart; Figure 3 This is a graph showing the stability index of time-domain rhythms; Figure 4 It is a graph combining confidence level and decision threshold. Detailed Implementation

[0025] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0027] This invention discloses a hotel energy optimization management method based on swarm intelligence, referring to... Figure 1 This includes steps S1-S5: S1: Obtain echo signals through millimeter-wave radar in the guest room, preprocess the echo signals to lock the target signal source and extract the time-domain displacement signal.

[0028] It should be noted that in the complex radio electromagnetic environment of hotel rooms, the echo signals received by millimeter-wave radar are often mixed with a large amount of static wall clutter, false reflections from non-target areas, and thermal noise from the circuit system. If target identification is performed directly based on the original echo signal, the extracted phase signal will have drastic jumps or a severely insufficient signal-to-noise ratio, making it impossible to accurately characterize the subtle breathing movements.

[0029] This step involves deploying a millimeter-wave radar in the center of the guest room ceiling. Multi-angle radiation coverage ensures no blind spots within the room. The radar transmits frequency-modulated continuous wave signals and receives echo signals. The echo signals are mixed and sampled to obtain an intermediate frequency signal, which is then processed using a fast Fourier transform along the fast time axis to generate a range Doppler heatmap. After filtering out static clutter using a background cancellation algorithm, the range cell with the strongest energy is identified as the target signal source to be analyzed. The phase signal of this range cell is extracted and then unwound and bandpass filtered to obtain a time-domain displacement signal.

[0030] S2: Construct a frequency domain energy concentration index based on time domain displacement signal.

[0031] It should be noted that the phase signals captured by millimeter-wave radar have extremely high similarity in energy performance, and the signal amplitude alone cannot distinguish between human respiration and environmental mechanical micro-movements. Human respiration is driven by the physiological center and has a single dominant frequency point in the frequency domain, while interference such as curtain swaying caused by air conditioning airflow is manifested as multiple frequency superposition or broadband distribution. Without frequency domain component purity analysis, the system is very likely to misjudge the high frequency of the interference source as the presence of human body.

[0032] Therefore, by evaluating the proportion of the dominant frequency energy, the system initially identifies the biological attributes of the target from the perspective of spectral structure. First, it performs high-resolution spectral analysis on the time-domain displacement signal to obtain the power spectrum density. Then, it calculates the ratio of the dominant peak energy to the total energy in the passband, obtaining the expression for the frequency domain energy concentration index: ; In the formula, It serves as a frequency domain energy concentration index, enabling a normalized assessment of the purity of signal frequency components; The power spectral density is obtained through high-resolution spectral analysis; The dominant frequency with the highest energy in the power spectrum represents the core frequency of the target's micro-motion; It is half of the main frequency bandwidth window, and its value is set according to the physiological bandwidth characteristics of human respiratory rate, which is used to define the spectral range of effective physiological signals. and These are the low-end and high-end cutoff frequencies of the bandpass filter, respectively, defining the total passband of the micro-frequency range of interest to the system. This is a numerical stability constant, taken as a minimum value, to prevent the denominator from being zero when the radar does not detect a moving target, thus ensuring computational stability. This represents the total energy within a specific bandwidth near the dominant frequency. This represents the total energy within the filter's passband, expressed as a stability constant. Ensure logical consistency.

[0033] This frequency domain energy concentration index calculates the proportion of local dominant frequency energy in the total energy to achieve a normalized assessment of signal focusing. A larger frequency domain energy concentration index indicates that the time-domain displacement signal energy is highly concentrated on a specific frequency, which is consistent with the single and stable single-frequency characteristics of human respiration, which is controlled by physiological functions. Conversely, a smaller frequency domain energy concentration index indicates that the time-domain displacement signal is interfered with by air conditioning airflow, such as swaying curtains, and its spectrum is characterized by multi-frequency spurious noise or broadband noise, with energy dispersed within the passband. This frequency domain energy concentration index achieves a preliminary decoupling of human signals from environmental mechanical noise from the frequency domain dimension.

[0034] Figure 2 This is a frequency domain energy concentration index analysis diagram. This diagram illustrates the principle by which the system identifies the biological attributes of a target through frequency domain energy distribution. Human respiratory signals are driven by physiological functions, and their power spectrum exhibits a single-frequency characteristic with extremely narrow peaks and significant peaks, with energy highly focused near the main frequency point. In contrast, mechanical interference in the environment is affected by airflow disturbances, and its energy exhibits a multi-peak stray or broadband noise distribution within the frequency band, with extremely low focus. By calculating the ratio of energy within the main frequency window to the total energy in the passband, the system can initially distinguish between living organisms and environmental noise from the perspective of frequency purity.

[0035] S3: Construct a time-domain rhythm stability index based on time-domain displacement signals.

[0036] It should be noted that some environmental interference sources may exhibit quasi-periodic pseudo-features in the frequency domain, and relying solely on frequency domain energy concentration indicators still carries the risk of false alarms; human physiological rhythms have a high degree of long-range conservation, while mechanical oscillations, due to the influence of airflow randomness, often experience severe jitter and drift in their vibration phase over time; without time-dimensional correlation verification, the system struggles to cope with feature confusion issues in non-stationary environments.

[0037] Therefore, this step utilizes the autocorrelation function to deeply analyze the periodic decay characteristics of the displacement signal. By constructing a time-domain rhythm stability index, it provides a crucial time-domain logical basis for distinguishing between physiological stability and mechanical randomness. The expression for the time-domain rhythm stability index is as follows: ; In the formula, It is a time-domain rhythm stability index used to evaluate the conservation and rhythm maintenance ability of micro-motion signals in the time dimension; The autocorrelation function of the extracted time-domain displacement signal; This is the zero-delay autocorrelation value of the signal, representing the total average power of the signal, used for normalization processing; The first autocorrelation function The correlation values ​​corresponding to each peak reflect the signal's response time during the delay period. Waveform similarity after the comparison; For the autocorrelation function, the th The delay time corresponding to each peak reflects the quasi-periodicity of the signal; The number of autocorrelation peaks is selected because an excessively small M value is difficult to fully reflect the multi-period rhythm characteristics, while an excessively large M value may introduce noise interference or non-dominant periodic components. Therefore, M is usually taken as a small positive integer, ranging from 2 to 5. In this embodiment, M=3. In other embodiments, the implementer can dynamically adjust this parameter according to the clarity of the signal period, the signal-to-noise ratio, and the intensity of environmental interference. It is a natural constant; The preset attenuation penalty coefficient is used to correct the natural attenuation of the autocorrelation peak as the delay time increases. However, if it is too small... The value is insufficient to effectively suppress the influence of unstable cycles, while an excessively large value... The value would excessively weaken the role of the multi-period cumulative criterion, therefore The value of is usually a positive number, ranging from 0.1 to 0.5. In this embodiment, it is set to... =0.3. In other embodiments, the implementer can dynamically adjust this parameter according to the intensity of environmental interference and the signal-to-noise ratio. It reflects the similarity between adjacent periodic signals; the exponential term This is an exponentially decaying penalty term used to measure the persistence of a rhythm.

[0038] This expression distinguishes micro-motion types based on autocorrelation decay characteristics. When the target is human respiration, due to the high stability of physiological rhythms, the peak value of the autocorrelation function decays slowly and is evenly distributed. The calculated value... The value is relatively high; if the target is a curtain or plant affected by wind, due to severe phase jitter, its autocorrelation peak will decay rapidly over time or become displaced, resulting in... The value decreased significantly; this indicator provides a time-domain basis for distinguishing between physiological regularity and environmental mechanical randomness.

[0039] Figure 3 This figure shows the time-domain rhythm stability index analysis, reflecting the differences in autocorrelation of micro-motion signals from different sources along the time axis. Micro-motion signals generated by human respiration have a strong rhythm maintenance ability; the peak value of their autocorrelation function remains high and evenly distributed as the delay time increases, demonstrating the long-term conservation of this type of physiological micro-motion characteristic. In contrast, mechanical disturbance signals in the environment are affected by random environmental forces and are often accompanied by severe phase jitter, causing their autocorrelation peak value to decay rapidly or become displaced with the delay time. The time-domain rhythm stability index constructed using the autocorrelation decay characteristics provides a reliable time-domain basis for distinguishing between physiologically regular living micro-motions and random environmental mechanical disturbances.

[0040] S4: The frequency domain energy concentration index and the time domain rhythm stability index are weighted and linearly fused, and the environmental noise bias parameter is introduced for correction to obtain the liveness confidence score that comprehensively reflects the strength of the target biological attributes.

[0041] It should be noted that in actual hotel room applications, due to differences in detection distance and real-time fluctuations in ambient noise, the reliability of single-dimensional features will exhibit dynamic shifts. If only a fixed-weight fusion strategy is used, the system is prone to missed detections due to weak signals in low signal-to-noise ratio environments. In strong interference scenarios, it is also prone to false recognitions due to feature confusion. Therefore, environmental background factors must be introduced as correction variables; otherwise, the robustness of the algorithm will not meet the needs of refined energy management.

[0042] Therefore, this step introduces an adaptive modulation mechanism based on the Sigmoid function to construct a multi-dimensional liveness confidence expression as follows: ; In the formula, For liveness confidence, it is a dimensionless probability value that reflects the strength of the target biological attribute; The signal-to-noise ratio (SNR) of the current range cell is used to measure the strength and quality of the received signal. It is an index of energy concentration in the frequency domain; As an index of temporal rhythm stability; These are weighting coefficients used to balance the contributions of the frequency domain energy concentration index and the time domain rhythm stability index to the final decision. For the preset weighting coefficients, satisfy Because the temporal rhythm stability index is more robust in distinguishing between physiological micro-movements and random mechanical perturbations, and can more effectively prevent feature confusion caused by environmental noise from leading to detection failure, a higher weight is given to the temporal rhythm stability index, i.e., it requires... In this embodiment, the following is set , In other embodiments, implementers may set weighting coefficients based on the actual implementation situation and complexity of the guest room environment; is the kurtosis adjustment factor of the Sigmoid function, which controls the sensitivity of signal-to-noise ratio changes to weight switching; g is the environmental noise floor bias parameter, used to define the initial signal-to-noise ratio threshold for the system's confidence in the frequency domain energy concentration index. Since a value of g that is too low is easily affected by false breathing signals in complex environments, leading to false alarms, while a value of g that is too high may suppress the sensitivity of recognizing real weak signals, the preset range of g is usually set to 15dB to 25dB. In this embodiment, g=20dB; in other embodiments, the implementer can dynamically adjust it according to the air conditioning layout in the guest room, the signal-to-noise ratio benchmark, and the intensity of environmental feedback. This constitutes a term for nonlinear modulation of the signal-to-noise ratio, used to achieve dynamic adjustment of feature weights; In reality, it's a sigmoid function, used as an adaptive gain factor for frequency domain energy concentration, and signal-to-noise ratio. The introduction of nonlinear weight adjustment reflects the difference in feature reliability under different signal intensities, and the weight coefficients... Achieve linear fusion of multi-dimensional indicators.

[0043] This expression achieves adaptive environmental decision-making through nonlinear mapping: when the signal-to-noise ratio... When the signal-to-noise ratio is low, the system automatically suppresses the influence of the frequency domain energy concentration index and relies more on the time domain rhythm stability index for logical judgment. When the value is high, the weight of the frequency domain energy concentration index increases accordingly.

[0044] S5: Compare the liveness confidence level with the preset judgment threshold. If the liveness confidence level is greater than the judgment threshold, determine that there is someone in the room and maintain the current energy mode. If the liveness confidence level is less than the judgment threshold, determine that there is no one in the room and execute the energy-saving program.

[0045] It should be noted that hotel room energy management needs to achieve a dynamic balance between guest satisfaction and energy saving depth. A single hard threshold cannot take into account the differences between the tranquil environment of a resort and the noisy environment of a commercial area. If the identification results are directly output without closed-loop interaction with historical environmental data, the system will not be able to self-correct identification biases, resulting in different energy-saving effects in different room types.

[0046] Therefore, T is set as the judgment threshold to control the sensitivity of the final decision. Since an excessively high T value is prone to missed detection under weak signals and affects guest satisfaction, while an excessively low T value is prone to false triggering due to noise interference in complex environments, T is usually taken as a value between 0.65 and 0.8. In this embodiment, T=0.75. In other embodiments, the implementer can dynamically adjust this parameter according to the type of guest room, background noise intensity and energy-saving goals. For example, a higher threshold can be set in a hotel in a noisy commercial area, while a lower threshold can be set in a resort hotel with high tranquility.

[0047] If the confidence level is in vivo If a stationary living entity is detected in the guest room, the system maintains the current comfort mode and sends a high-confidence occupancy status message to the swarm intelligence center to prevent accidental shutdown; if If the system determines that there are no living beings or only environmental interference in the guest room, it will execute a false occupancy identification process, gradually reducing the air conditioning load or cutting off unnecessary power. In addition, the determination result will be uploaded to the cloud-based swarm intelligence system as a kind of perception feedback. The system will dynamically adjust the environmental noise bias parameter and judgment threshold in the algorithm for that guest room based on the historical false alarm rate, so as to achieve adaptive evolution of the algorithm parameters.

[0048] Figure 4 This diagram illustrates the system decision-making process when a guest room transitions from an occupied state to a complex interference environment, using a comprehensive confidence level and a judgment threshold. The system dynamically compares the real-time calculated comprehensive liveness confidence level with a preset judgment threshold: when a signal matching physiological characteristics is detected, causing the confidence level to exceed the judgment threshold, the system determines that the guest is present and maintains power supply; when faced with interference from false targets such as swaying curtains, due to the low frequency domain energy concentration index and time domain rhythm stability index, the comprehensive confidence level is suppressed below the judgment threshold, and the system accurately identifies the guest as unoccupied and executes energy-saving procedures.

[0049] The present invention also discloses a hotel energy optimization management system based on swarm intelligence, including a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, the hotel energy optimization management method based on swarm intelligence according to the present invention is implemented.

[0050] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.

Claims

1. A hotel energy optimization management method based on swarm intelligence, characterized in that, include: The echo signal is acquired by a millimeter-wave radar deployed in the guest room. The echo signal is preprocessed to lock the target signal source and extract the time-domain displacement signal. The frequency domain energy concentration index is obtained based on the time-domain displacement signal. The frequency domain energy concentration index is the ratio of the total energy in a specific bandwidth near the main frequency point in the power spectrum of the time-domain displacement signal to the total energy in the passband of the filter. It is used to evaluate the purity of the signal's frequency components. The time-domain rhythm stability index is also obtained based on the time-domain displacement signal. The time-domain rhythm stability index is determined by the ratio of the correlation values ​​of multiple peaks in the autocorrelation function of the time-domain displacement signal to its zero-delay autocorrelation value. It is used to evaluate the signal's ability to maintain rhythm in the time dimension. The frequency domain energy concentration index and the time domain rhythm stability index are weighted and linearly fused based on a weighted linear model, and an environmental noise bias parameter is introduced for correction to obtain a liveness confidence score that comprehensively reflects the strength of the target biological attributes. The liveness confidence score is compared with a preset judgment threshold. If the liveness confidence score is greater than the judgment threshold, it is determined that there is someone in the room and the current energy mode is maintained. If the liveness confidence score is less than the judgment threshold, it is determined that there is no one in the room and the energy-saving program is executed.

2. The hotel energy optimization management method based on swarm intelligence according to claim 1, characterized in that, The millimeter-wave radar is deployed in the center of the guest room ceiling, ensuring no blind spots in the guest room through multi-angle radiation coverage.

3. The hotel energy optimization management method based on swarm intelligence according to claim 1, characterized in that, The preprocessing of the echo signal includes mixing and sampling the echo signal to obtain an intermediate frequency signal, followed by fast Fourier transform processing along the fast time axis to generate a range Doppler heatmap; after filtering out static clutter based on the background cancellation algorithm, the range cell with the strongest energy is locked as the target signal source to be analyzed, the phase signal of the range cell is extracted, and the phase signal of the range cell is unwound and bandpass filtered to obtain a time-domain displacement signal reflecting the micro-motion characteristics of the target.

4. The hotel energy optimization management method based on swarm intelligence according to claim 1, characterized in that, The expression for the frequency domain energy concentration index is: ； In the formula, It is an index of energy concentration in the frequency domain; The power spectral density is obtained through high-resolution spectral analysis; This is the dominant frequency point with the highest energy in the power spectrum; Half of the main frequency bandwidth window; and These are the low-end cutoff frequency and the high-end cutoff frequency of the bandpass filter, respectively. It is the numerical stability constant; This represents the total energy within a specific bandwidth near the dominant frequency. This represents the total energy within the filter's passband.

5. The hotel energy optimization management method based on swarm intelligence according to claim 1, characterized in that, The expression for the time-domain rhythm stability index is: ； In the formula, As an index of temporal rhythm stability; The autocorrelation function of the extracted time-domain displacement signal; This is the zero-delay autocorrelation value of the signal; The first autocorrelation function The correlation values ​​corresponding to each peak; For the autocorrelation function, the th The delay time corresponding to each peak; The number of selected autocorrelation peaks; It is a natural constant; This is the preset attenuation penalty coefficient.

6. The hotel energy optimization management method based on swarm intelligence according to claim 1, characterized in that, The expression for the liveness confidence level is: ； In the formula, For liveness confidence; It is an index of energy concentration in the frequency domain; The signal-to-noise ratio of the current distance cell; This is the environmental noise floor bias parameter; Preset weighting coefficients; This is the kurtosis adjustment factor for the Sigmoid function; It is an index for the stability of time-domain rhythms.

7. The hotel energy optimization management method based on swarm intelligence according to claim 6, characterized in that, The preset weighting coefficient .

8. The hotel energy optimization management method based on swarm intelligence according to claim 1, characterized in that, The energy-saving procedures include gradually reducing the air conditioning load or cutting off unnecessary power sources.

9. The hotel energy optimization management method based on swarm intelligence according to claim 1, characterized in that, The method also includes uploading the determination result as a kind of perception feedback to a cloud-based swarm intelligence system. The system will dynamically adjust the environmental noise bias parameter and determination threshold in the algorithm for the room based on the historical false alarm rate of the room.

10. A hotel energy optimization management system based on swarm intelligence, characterized in that: include: A processor and a memory, the memory storing computer program instructions that, when executed by the processor, implement the hotel energy optimization management method based on swarm intelligence according to any one of claims 1-9.

Images