A method for intelligent lighting energy-saving control through target tracking

Abstract

The application discloses a method for realizing intelligent lighting energy-saving control through target tracking, which comprises the following steps: establishing a self-organizing single-point network by using wireless sensors, so that each sensor node in the wireless sensor network is bound with a nearby lighting source; performing particle filtering analysis on the target by taking the node where the target is located as the current time following node, predicting the target position and speed, and taking the prediction as the basis for waking up the next time following node; calculating the distance between the predicted target position and the current real position, and judging whether the predicted target position falls within the wake-up range of the current following node; if the predicted position falls outside the wake-up range of the current following node, then adjusting the sampling time according to the ratio of the node wake-up radius to the predicted speed; performing particle filtering calculation again on the adjusted sampling time, re-predicting the current time target position and speed; and waking up the node with the largest probability in the filtering result near the next time predicted position and taking the node as the following node at the next time.

Description

Technical Field

[0001] This invention relates to the fields of wireless sensor networks and particle filter positioning technology, and in particular to a method for achieving intelligent lighting energy-saving control through target tracking. Background Technology

[0002] As people's living standards continue to improve, their demands for working and living environments are also increasing, as are their requirements for lighting systems. Lighting accounts for a significant proportion of total energy consumption, making energy conservation and improved lighting quality urgent priorities. Lighting electricity consumption, as a major part of electricity consumption, already accounts for about 10% of total electricity consumption, and this will continue to increase with the rapid development of my country's national economy and the continuous improvement of people's living standards. Previously, due to cost and immature technology, lighting alone accounted for 33% of building energy consumption (air conditioning accounted for 50%, and other uses accounted for 17%). With my country's sustainable development strategy, contributing to achieving carbon peaking before 2030 and carbon neutrality before 2060, optimizing lighting control is key to energy conservation and emission reduction; therefore, intelligent lighting control systems have been widely applied in office buildings.

[0003] Technical solutions of existing technologies

[0004] Currently, infrared motion detection technology is widely used. When someone in the lighting environment needs light, the lighting control system turns on the light source; otherwise, if no one needs light, the light source is turned off, achieving energy-saving effects. Another technology is voice control, which determines whether the lighting source needs to be turned on or off by detecting the volume of sound.

[0005] The shortcomings of existing technologies:

[0006] Existing infrared or sound detection methods can only determine whether something is present or not, and the results are static and cannot be known in advance.

[0007] When the target temperature is close to the ambient temperature, it is particularly susceptible to environmental interference. The sensitivity will decrease with factors such as ambient temperature and noise, and infrared detection is prone to failure. Sound detection also has similar problems.

[0008] There are many algorithms for predicting target trajectories, but they are computationally intensive and require high-performance computing equipment. When the number of nodes is large, the engineering cost is high.

[0009] Infrared detection has poor penetration, and there must be no obstruction between the target and the detection device, which places high demands on installation and deployment.

[0010] Without target recognition and unable to distinguish between human and animal targets, it is easy to turn on lighting for irrelevant targets, resulting in a waste of resources. Summary of the Invention

[0011] To address the aforementioned technical problems, the present invention aims to provide a method for intelligent lighting energy-saving control through target tracking. This method is based on wireless sensors and particle filter positioning, and uses a specific target trajectory prediction algorithm to link the lighting switch according to the target's movement trajectory, thereby achieving precise lighting control and energy saving.

[0012] The objective of this invention is achieved through the following technical solution:

[0013] A method for achieving intelligent lighting energy-saving control through target tracking includes:

[0014] Step A uses wireless sensors to establish a self-organizing single-point network, so that each sensor node in the wireless sensor network is bound to a nearby lighting source. The wake-up of a node means the lighting source is turned on, and the sleep of a node means the lighting source is turned off.

[0015] Step B uses the node where the target is located as the current follower node to perform particle filtering analysis on the target, predict the target's position and velocity, and use this as the basis for waking up the follower node at the next moment.

[0016] Step C calculates the distance between the predicted target location and the current actual location, and determines whether the predicted target location falls within the wake-up range of the current following node;

[0017] If the predicted position in step D falls outside the wake-up range of the current following node, the sampling time is adjusted according to the ratio of the node wake-up radius to the prediction speed.

[0018] Step E involves recalculating the particle filter for the adjusted sampling time to re-predict the target position and velocity at the current moment.

[0019] Step F wakes up the node with the highest probability in the filtering results near the predicted position at the next time step and uses it as the follow node for the next time step.

[0020] Compared with the prior art, one or more embodiments of the present invention may have the following advantages:

[0021] It can accurately detect the location of the target and predict its movement, thereby achieving more intelligent and precise linkage control of lighting sources;

[0022] Wireless network sensors rely on electromagnetic signals for communication and are almost unaffected by interference from the natural environment.

[0023] The prediction location algorithm is simple and effective, and can run on low-computing-power embedded CPUs. Attached Figure Description

[0024] Figure 1This is a flowchart illustrating a method for achieving intelligent lighting energy-saving control through target tracking.

[0025] Figure 2 This is a flowchart illustrating the specific implementation of intelligent lighting energy-saving control through target tracking.

[0026] Figure 3 This is a schematic diagram of target tracking. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be described in further detail below with reference to the embodiments and accompanying drawings.

[0028] like Figure 1 The diagram shows a method for achieving intelligent lighting energy-saving control through target tracking, including the following steps:

[0029] Step 101: Establish a self-organizing point network using wireless sensors, so that each sensor node in the wireless sensor network is bound to a nearby lighting source. The wake-up of a node means the lighting source is turned on, and the sleep of a node means the lighting source is turned off.

[0030] Step 102 uses the node where the target is located as the current follower node to perform particle filtering analysis on the target, predict the target's position and velocity, and use this as the basis for waking up the follower node at the next moment.

[0031] Step 103: Calculate the distance between the predicted target location and the current actual location, and determine whether the predicted target location falls within the wake-up range of the current following node;

[0032] If the predicted position falls outside the wake-up range of the current following node in step 104, the sampling time is adjusted according to the ratio of the node wake-up radius to the prediction speed.

[0033] Step 105 recalculates the particle filter for the adjusted sampling time and re-predicts the target position and velocity at the current moment;

[0034] Step 106 wakes up the node with the highest probability in the filtering results near the predicted position at the next moment and uses it as the follow node for the next moment.

[0035] like Figure 2 As shown, the method for achieving intelligent lighting energy-saving control by target tracking specifically includes the following steps:

[0036] Step 201: Initialize the wireless sensor network. Wireless sensor nodes in the detection area are bound to nearby lighting sources, with their distribution mirroring the lighting sources. All nodes have the same communication and wake-up capabilities; their wake-up and sleep states are linked to the bound lighting source. A node waking up turns on the lighting source, and a node sleeping turns off the lighting source. Sampling time t kThe initial value is 0. Only a few nodes located at the building entrances and exits are in probe mode, while most other nodes are in sleep mode. All lighting sources are off.

[0037] Step 202: When a target is detected within the detection area of ​​the wireless sensor network, select the node k closest to the target and record time t. k And let k be the current time t k The follower node of the tracked target. This node is in a wake-up state, and the lighting source bound to it is turned on.

[0038] Step 203: Node k is responsible for obtaining the measured value and updating the target position to obtain a local estimate, while performing particle filtering prediction t. k+1 The target's state at any given time, including its predicted position and velocity, serves as the basis for waking up the sensor nodes.

[0039] Step 204 calculates the distance between the predicted target location and the sensor node wake-up radius. Based on this distance, the relationship between the predicted target location and the node wake-up range is determined. There are two cases: the predicted target location falls within the node wake-up range and the predicted target location falls outside the node wake-up range. Different processing is performed for these two different cases.

[0040] Step 205: When the predicted target location falls within the node wake-up range, directly select the node closest to the target as t. k+1 At time k+1, when the target's predicted position falls outside the node wake-up range, the sensor nodes that are woken up near the predicted position may exceed the effective wake-up range of the target following node k at the current time, causing the target to lose track. In this case, it is necessary to combine the prediction speed and the wake-up radius to calculate the sampling time that needs to be re-predicted and perform particle filtering to predict the target state. Then, select the head node k+1 again based on the new predicted position.

[0041] Step 206 wakes up the selected t k+1 At time k+1, the packet information sends the target state estimate from head node k to head node k+1, and simultaneously puts head node k into a sleep state. At this time, the lighting source bound to node k is turned off, and the lighting source associated with node k+1 is turned on.

[0042] Step 207: The sampling time increments by 1.

[0043] Step 208: Repeat steps 201 to 207 above until the target leaves the detection area.

[0044] See Figure 3If the distance between the predicted position and the current position at any given moment exceeds the node's wake-up range, the inability to effectively wake up the node will lead to target loss of tracking. This demonstrates the significant impact of target state prediction on the target loss rate. Approximating the target's predicted position as the node to be woken up at the next moment, R... s Let t be the node wake-up radius. k The current position of the activity target A is l k (x k y k ), t k+1 Predicted position B at time is Let Δ be the distance between them, then we have:

[0045]

[0046]

[0047] Substituting (2) into (1), we have:

[0048] Δ=g(T,v k+1 (3)

[0049] Δ and T, v k They are directly proportional. Δ≤R s The predicted location falls within the wake-up range of the current node; Δ > R s The predicted location falls outside the current node's wake-up range. For Δ > R... s The situation, through joint R s and Adjust the sampling time to T′ (see Figure 1 ):

[0050]

[0051] As can be seen from (4), T′ is The decreasing function. Based on T′, the particle filter prediction is performed again, and the predicted position B′ is obtained as... While ensuring network connectivity and full coverage of the target area, if the predicted location The target is lost if it still falls outside the wake-up range of the current node. Let P k+1 For t k+1 If the target loses tracking rate at any given moment, then:

[0052]

[0053]

[0054] When the sampling time is constant, P k+1 yes The rate of target loss is increasing as the target's speed increases; when the target's speed increases, the sampling time should be reduced accordingly to decrease the target loss rate.

[0055] While the embodiments disclosed in this invention are as described above, the content is merely for the purpose of facilitating understanding of the invention and is not intended to limit the invention. Any person skilled in the art to which this invention pertains may make any modifications and variations in form and detail of the implementation without departing from the spirit and scope disclosed herein; however, the scope of patent protection for this invention shall still be determined by the scope defined in the appended claims.

Claims

1. A method for achieving intelligent lighting energy-saving control through target tracking, characterized in that, The method includes the following steps: Step A uses wireless sensors to establish a self-organizing single-point network, so that each sensor node in the wireless sensor network is bound to a nearby lighting source. The wake-up of a node means the lighting source is turned on, and the sleep of a node means the lighting source is turned off. Step B uses the node where the target is located as the current follower node to perform particle filtering analysis on the target, predict the target's position and velocity, and use this as the basis for waking up the follower node at the next moment. Step C calculates the distance between the predicted target location and the current actual location, and determines whether the predicted target location falls within the wake-up range of the current following node; If the predicted position in step D falls outside the wake-up range of the current following node, the sampling time is adjusted according to the ratio of the node wake-up radius to the prediction speed. Step E involves recalculating the particle filter for the adjusted sampling time to re-predict the target position and velocity at the current moment. Step F wakes up the node with the highest probability in the filtering results near the predicted position at the next moment and uses it as the follow node at the next moment; In step B, the wireless sensor network detects a target within the detection area, selects the node k closest to the target, and records time t. k And let k be the current time t k The tracking node is in a wake-up state, and the illumination source bound to it is turned on. Node k is responsible for acquiring measurements and updating the target position to obtain a local estimate, while simultaneously performing particle filtering prediction t. k+1 The target state at any time, including the predicted position and velocity of the target, is used as the basis for waking up the next sensor node; Step C specifically includes: calculating the distance between the predicted target location and the sensor node wake-up radius, determining the relationship between the predicted target location and the node wake-up range based on the distance value, and identifying two cases: the predicted target location falls within the node wake-up range and the predicted target location falls outside the node wake-up range, and then handling each case differently. When the predicted target location falls within the node wake-up range, the node closest to the target is directly selected as t. k+1 Time header node k+1; When the predicted target position falls outside the node wake-up range, the sensor nodes that are woken up near the predicted position may exceed the effective wake-up range of the target following node k at the current moment, causing the target to lose tracking. In this case, it is necessary to combine the prediction speed and the wake-up radius to calculate the sampling time that needs to be re-predicted and perform particle filtering to predict the target state. Then, the head node k+1 is selected based on the new predicted position.

2. The method for intelligent lighting energy-saving control by target tracking as described in claim 1, characterized in that, Step A specifically includes initializing the wireless sensor network and binding the wireless sensor nodes in the detection area of ​​the wireless sensor network to nearby lighting sources, so that the distribution of sensor nodes is consistent with the lighting coverage.

3. The method for achieving intelligent lighting energy-saving control through target tracking as described in claim 2, characterized in that, All sensor nodes have the same communication and wake-up capabilities. Their wake-up and sleep modes are linked to the attached lighting source. When a node wakes up, the lighting source is turned on; when a node goes into sleep mode, the lighting source is turned off.

4. The method for achieving intelligent lighting energy-saving control through target tracking as described in claim 1, characterized in that, Wake up the selected t k+1 At time k+1, the packaged information sends the target state estimate from head node k to head node k+1, and at the same time, head node k enters a sleep state; at this time, the lighting source bound to node k is turned off, and the lighting source associated with node k+1 is turned on.

Images