WO2019214016A1 - 基于LoRa技术的多功能LED智能路灯系统 - Google Patents

基于LoRa技术的多功能LED智能路灯系统 Download PDF

Info

Publication number
WO2019214016A1
WO2019214016A1 PCT/CN2018/092617 CN2018092617W WO2019214016A1 WO 2019214016 A1 WO2019214016 A1 WO 2019214016A1 CN 2018092617 W CN2018092617 W CN 2018092617W WO 2019214016 A1 WO2019214016 A1 WO 2019214016A1
Authority
WO
WIPO (PCT)
Prior art keywords
street lamp
main processor
lora
layer
fuzzy
Prior art date
Application number
PCT/CN2018/092617
Other languages
English (en)
French (fr)
Inventor
彭力
邢睿智
方超
边栋
董君怡
董子豪
谢林柏
Original Assignee
江南大学
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 江南大学 filed Critical 江南大学
Publication of WO2019214016A1 publication Critical patent/WO2019214016A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H05—ELECTRIC TECHNIQUES NOT OTHERWISE PROVIDED FOR
    • H05B—ELECTRIC HEATING; ELECTRIC LIGHT SOURCES NOT OTHERWISE PROVIDED FOR; CIRCUIT ARRANGEMENTS FOR ELECTRIC LIGHT SOURCES, IN GENERAL
    • H05B45/00—Circuit arrangements for operating light-emitting diodes [LED]
    • H05B45/10—Controlling the intensity of the light
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02B—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO BUILDINGS, e.g. HOUSING, HOUSE APPLIANCES OR RELATED END-USER APPLICATIONS
    • Y02B20/00—Energy efficient lighting technologies, e.g. halogen lamps or gas discharge lamps
    • Y02B20/30—Semiconductor lamps, e.g. solid state lamps [SSL] light emitting diodes [LED] or organic LED [OLED]
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02B—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO BUILDINGS, e.g. HOUSING, HOUSE APPLIANCES OR RELATED END-USER APPLICATIONS
    • Y02B20/00—Energy efficient lighting technologies, e.g. halogen lamps or gas discharge lamps
    • Y02B20/40—Control techniques providing energy savings, e.g. smart controller or presence detection

Definitions

  • the invention relates to a street lamp management system, in particular to a multifunctional LED intelligent street lamp system based on LoRa technology.
  • the electric light source used in the field of road lighting is still dominated by high-pressure sodium lamps, accounting for about 70% of all light sources.
  • Traditional street lighting generally has manual, light control, clock control and other control forms, which are susceptible to seasons and weather. The time and brightness of the switch lights cannot be adjusted according to actual conditions. When the season changes or the weather is abnormal, the street lights must be manually operated one by one. This simple control method is seriously inflexible; in the night when no one is walking or few people are walking, the street lights are still working normally, causing great energy waste; without street lamp status monitoring function, real-time monitoring of operating conditions and collection of operational data is not possible. The fault mainly relies on manual inspection, which requires huge manpower and time cost, and the reliability and accuracy are poor, which also brings inconvenience to the citizens' lives.
  • a multifunctional LED intelligent street lamp system based on LoRa technology comprising:
  • the detection module detects a change in illumination brightness of the environment and an activity state of the pedestrian or the vehicle;
  • main processor controlling the street lamp according to the data detected by the detecting module
  • the LoRa wireless communication module is connected to the GPRS module and the main processor, and sends an instruction received from the web end to the main processor to implement remote control of the street lamp, and the data detected by the detection module is transmitted to Server;
  • the system terminal receives the data detected by the measurement module transmitted by the LoRa wireless communication module, and displays the measured data on the webpage through data processing, and provides a view of the street light state, the street light switch, the remote adjustment brightness, and the statistical analysis function through the web interface. Visualization of consumption and fault monitoring.
  • the above-mentioned multi-function LED intelligent street lamp system based on LoRa technology embeds intelligent control modules such as sensing and wireless transmission on the basis of the traditional lighting function of street lamps, so that the street lamps have automatic brightness adjustment, remote adjustment brightness, self-reporting of faults, and statistical analysis. Energy consumption and other functions.
  • the detection module includes a photo sensor, a sound sensor, and an infrared sensor.
  • the primary processor is an STM32F030 microprocessor.
  • the algorithm for controlling the street lamp by the main processor according to the data detected by the detecting module is a fuzzy neural network algorithm with compensation.
  • the compensated fuzzy neural network algorithm includes an input layer, a fuzzification layer, a compensated fuzzy inference layer, and an anti-fuzzification layer;
  • Each node of the input layer is directly connected to an input vector, and the input quantity includes time, location, illumination intensity, pedestrian, vehicle, etc.; each node of the fuzzification layer represents a linguistic variable value, and its function is to calculate each The input vector belongs to the membership function of the fuzzy set of each linguistic variable value; each node of the compensated fuzzy inference layer represents a fuzzy rule, and its function is to match the fuzzy rule, calculate the applicability of each rule, and perform compensation operation;
  • the defuzzification layer gives the output turn-on information.
  • FIG. 1 is a schematic structural diagram of a fuzzy neural network algorithm with compensation in a multi-function LED intelligent street lamp system based on LoRa technology according to an embodiment of the present application.
  • a multifunctional LED intelligent street lamp system based on LoRa technology comprising:
  • the detection module detects a change in illumination brightness of the environment and an activity state of the pedestrian or the vehicle;
  • main processor controlling the street lamp according to the data detected by the detecting module
  • the LoRa wireless communication module is connected to the GPRS module and the main processor, and sends an instruction received from the web end to the main processor to implement remote control of the street lamp, and the data detected by the detection module is transmitted to Server;
  • the system terminal receives the data detected by the measurement module transmitted by the LoRa wireless communication module, and displays the measured data on the webpage through data processing, and provides a view of the street light state, the street light switch, the remote adjustment brightness, and the statistical analysis function through the web interface. Visualization of consumption and fault monitoring.
  • the above-mentioned multi-function LED intelligent street lamp system based on LoRa technology embeds intelligent control modules such as sensing and wireless transmission on the basis of the traditional lighting function of street lamps, so that the street lamps have automatic brightness adjustment, remote adjustment brightness, self-reporting of faults, and statistical analysis. Energy consumption and other functions.
  • the detection module includes a photo sensor, a sound sensor, and an infrared sensor.
  • the primary processor is an STM32F030 microprocessor.
  • the sensor module (detection module) is mainly composed of a photosensitive sensor, a sound sensor and an infrared sensor. The three comprehensive judgments can effectively detect the passage of people and vehicles, and reduce the impact of interference signals on the system.
  • the main processor used in this system in the intelligent control module is the STM32F030 microprocessor. Low power consumption, high performance, strong practicability, high cost performance, etc., suitable for applications.
  • LoRa wireless communication module In this system, the LoRa module is connected with the GPRS module and the single chip microcomputer, and sends the command received from the web end to the STM32 controller to realize remote control of the street lamp, and can transmit the state data such as the road to the server. To monitor and analyze the status of street lights.
  • GPRS module In this system, because the coverage of intelligent street lamps is very wide, and communication with the control terminal is needed, the GPRS communication technology is adopted, which increases the reliability of the network and facilitates data transmission.
  • This system includes a front-end web system and a back-end information processing system.
  • the front-end web system exposes the system administrator in the form of a web page to control the smart street lights associated with the system and the monitoring of the status of the street lights.
  • the back-end information processing system processes, sends, saves, and the like information sent by the remote hardware and information sent by the front-end web page.
  • the algorithm for controlling the street lamp by the main processor according to the data detected by the detecting module is a fuzzy neural network algorithm with compensation.
  • the compensated fuzzy neural network algorithm includes an input layer, a fuzzification layer, a compensated fuzzy inference layer, and an anti-fuzzification layer;
  • Each node of the input layer is directly connected to an input vector, and the input quantity includes time, location, illumination intensity, pedestrian, vehicle, etc.; each node of the fuzzification layer represents a linguistic variable value, and its function is to calculate each The input vector belongs to the membership function of the fuzzy set of each linguistic variable value; each node of the compensated fuzzy inference layer represents a fuzzy rule, and its function is to match the fuzzy rule, calculate the applicability of each rule, and perform compensation operation;
  • the defuzzification layer gives the output turn-on information.
  • the layer and layer are constructed according to the linguistic variables of the fuzzy logic system, the compensated fuzzy inference method, and the anti-fuzzification function.
  • the structure is shown in Figure 1.
  • the domain of input space x is The domain of the output space y is Then the fuzzy segmentation of the input space x can be performed according to the following principle: the subspace which changes sharply with y changes with x is subdivided into x, and the subspace which changes gently with y varies with x. Specifically, it is possible to analyze the variation of y with x, and use x to change the extreme point and the number of inflection points of y (the two endpoints are neither extreme points nor inflection points) in order to determine y in the sub-space. The intensity of the change in space, the more sub-spaces of extreme points and inflection points, the more fuzzy segments are divided for x, and vice versa.
  • fuzzy segmentation point may not be the mean point in the input domain
  • Gaussian and bell-shaped functions are used as membership functions.
  • the selection of the initial values of each parameter in the network is very demanding, otherwise the network training will fall into a chaotic state and the output error will not be achieved.
  • Accuracy requirements, so the trigonometric function is chosen as the membership function.
  • M is a large number, which must extend the calculation speed of the model. It can be seen that if the value of M can be reduced, it can be greatly improved. The calculation speed of the model.
  • the rules are valid.
  • the output of the second layer is Then there is
  • the output value of the corresponding node in the third layer with the input end is zero, and the node value in the third layer may not be calculated, that is, the node is omitted.
  • the location size of the third layer of these nodes can be expressed by the following formula:
  • the nodes in the third layer proposed in the invention are much less than the conventional structure, thereby reducing the amount of calculation and increasing the calculation speed.
  • the network of the invention is also a multi-layer feedforward network in nature, so the learning algorithm for adjusting parameters can be designed by using the error back propagation method of the feedforward network.
  • the error cost function is set to:
  • an output fuzzy subset B' can be generated in the output domain V. Fuzzy reasoning uses the algebraic product ( ⁇ ) operation, then the fuzzy set B' on V derived by the fuzzy inference rule is
  • n is the dimension of the input vector
  • ⁇ is the degree of compensation, ⁇ [0,1].
  • the system consists of a triangular membership function, compensated fuzzy inference, and an improved center-of-gravity defuzzifier.
  • the objective function is:
  • the gradient descent method of the dynamic adjustment step is used to train the input and output parameters of the system and the compensation degree of the compensation operation.
  • the corresponding iteration formula is:
  • control scheme of the specific intelligent street lamp system is as follows:
  • the Web first selects the system working mode, including the automatic control mode and the real-time control mode, and then sends the command to the LoRa module through the GPRS module through the cloud server.
  • the LoRa module sends the command to the STM32 controller on each street lamp node, and the controller analyzes the command and Control the status of the street light.
  • the controller controls the street light node in full accordance with the server instruction, realizing the real-time control of the street light node by the server.
  • the controller controls the street light node according to the preset control strategy, and turns on the street light in the case that the daytime illumination is not strong, and the street light remains on before the early morning of the night, and after the morning and when the person passes, the vehicle turns on. Street light, otherwise off.
  • the conditions for specifically turning on the street light are determined according to a data fusion algorithm.
  • the system uses low-power LoRa technology, STM32 as the core data processing chip, using photosensitive, sound, infrared and current sensors to detect changes in the ambient light and the activity of pedestrians and vehicles.
  • STM32 chip automatically adjusts the brightness of the street light according to the information of the acquisition sensor; wireless data transmission through LoRa, the data is transmitted to the system terminal, and the measured data is displayed on the webpage through data processing, and the web interface provides viewing status of the street light, street light switch, remote Visualize the brightness, statistical analysis of energy consumption, and fault monitoring.
  • the intelligent street lamp system based on LoRa technology will greatly reduce power resources, improve lighting management efficiency, reduce maintenance costs, and provide convenient information and services, which is of great significance for building smart cities.
  • the application of LoRa technology in the field of intelligent street lamps is basically in a blank position and has broad market prospects.
  • the invention combines the Internet of Things technology with street lighting, and designs a multi-function LED intelligent street lamp system based on LoRa technology for the background of intelligent street lighting. Based on the traditional lighting function of the street lamp, the sensing and wireless are embedded.
  • the intelligent control module such as transmission enables the street lamp to automatically adjust the brightness, adjust the brightness remotely, report the fault automatically, and analyze the energy consumption.
  • the invention mainly solves the problem of single function and lack of intelligence in the traditional street lamp illumination.
  • the traditional street lamp illumination is susceptible to weather conditions, the control mode lacks flexibility, the energy waste is large, the manual inspection consumes manpower and time, and the operation state cannot be monitored in real time. And collecting operational data, so it is necessary to study the multifunctional intelligent street lighting system, using a new neural network fusion technology to make up for the deficiencies of the traditional street lamp singulation function.

Landscapes

  • Circuit Arrangement For Electric Light Sources In General (AREA)

Abstract

本发明涉及一种基于LoRa技术的多功能LED智能路灯系统，包括：检测模块，所述检测模块检测环境的光照亮度变化以及行人、车辆的活动状况；主处理器，所述主处理器根据检测模块检测到的数据控制路灯；LoRa无线通信模块，所述LoRa无线通信模块与GPRS模块和主处理器相连，把从web端接收到的指令发送到主处理器，实现对路灯的远程控制，同时检测模块检测到的数据传送至服务器；以及系统终端。上述基于LoRa技术的多功能LED智能路灯系统，在路灯传统的照明作用基础上，嵌入了传感、无线传输等智能控制模块，使路灯具有自动调节亮度、远程调节亮度、故障自主上报、统计分析能耗等功能。

Description

基于LoRa技术的多功能LED智能路灯系统 技术领域
本发明涉及路灯管理系统，特别是涉及基于LoRa技术的多功能LED智能路灯系统。
背景技术
目前，在道路照明领域使用的电光源仍以高压钠灯为主，约占所有光源的70％。传统的路灯照明一般有手动、光控、钟控等控制形式，易受季节、天气影响，不能根据实际情况调节开关灯时间及亮度，在季节变化或者天气异常时，必须人工逐个操作路灯，这种简单的控制方式严重缺乏灵活性；在无人行走或少人行走的夜间，路灯仍然正常工作，造成极大的能源浪费；不具备路灯状态监测功能，不能实时监控运行状况和收集运行数据，故障主要靠人工巡检，需要耗费巨大的人力和时间成本，而且可靠性、准确性差，也为市民生活带来不便。
发明内容
基于此，有必要针对上述技术问题，提供一种基于LoRa技术的多功能LED智能路灯系统。
一种基于LoRa技术的多功能LED智能路灯系统，包括：
检测模块，所述检测模块检测环境的光照亮度变化以及行人、车辆的活动状况；
主处理器，所述主处理器根据检测模块检测到的数据控制路灯；
LoRa无线通信模块，所述LoRa无线通信模块与GPRS模块和主处理器相连，把从web端接收到的指令发送到主处理器，实现对路灯的远程控制，同时检测模块检测到的数据传送至服务器；以及
系统终端，所述系统终端接收通过LoRa无线通信模块传输的测模块检测到的数据，经数据处理将所测数据在网页显示，Web界面提供查看路灯状态、路 灯开关、远程调节亮度、统计分析能耗和故障监控的可视化操作。
上述基于LoRa技术的多功能LED智能路灯系统，在路灯传统的照明作用基础上，嵌入了传感、无线传输等智能控制模块，使路灯具有自动调节亮度、远程调节亮度、故障自主上报、统计分析能耗等功能。
在另外的一个实施例中，所述检测模块包括光敏传感器、声音传感器和红外传感器。
在另外的一个实施例中，所述主处理器是STM32F030微处理器。
在另外的一个实施例中，所述主处理器根据检测模块检测到的数据控制路灯的算法是带补偿的模糊神经网络算法。
在另外的一个实施例中，所述带补偿的模糊神经网络算法包括输入层、模糊化层、补偿模糊推理层和反模糊化层；
所述输入层的各个节点直接与输入向量相连接，输入量包括了时间、地点、光照强度、行人、车辆等；所述模糊化层的每一个节点代表一个语言变量值，其作用是计算各输入向量属于各语言变量值模糊集合的隶属函数；所述补偿模糊推理层的每一个节点代表一条模糊规则，其作用是匹配模糊规则、计算出每条规则的适用度，并进行补偿运算；所述反模糊化层给出输出量开灯信息。
附图说明
图1为本申请实施例提供的一种基于LoRa技术的多功能LED智能路灯系统中带补偿的模糊神经网络算法的结构示意图。
具体实施方式
为了使本发明的目的、技术方案及优点更加清楚明白，以下结合附图及实施例，对本发明进行进一步详细说明。应当理解，此处所描述的具体实施例仅仅用以解释本发明，并不用于限定本发明。
一种基于LoRa技术的多功能LED智能路灯系统，包括：
检测模块，所述检测模块检测环境的光照亮度变化以及行人、车辆的活动状况；
主处理器，所述主处理器根据检测模块检测到的数据控制路灯；
LoRa无线通信模块，所述LoRa无线通信模块与GPRS模块和主处理器相连，把从web端接收到的指令发送到主处理器，实现对路灯的远程控制，同时检测模块检测到的数据传送至服务器；以及
系统终端，所述系统终端接收通过LoRa无线通信模块传输的测模块检测到的数据，经数据处理将所测数据在网页显示，Web界面提供查看路灯状态、路灯开关、远程调节亮度、统计分析能耗和故障监控的可视化操作。
上述基于LoRa技术的多功能LED智能路灯系统，在路灯传统的照明作用基础上，嵌入了传感、无线传输等智能控制模块，使路灯具有自动调节亮度、远程调节亮度、故障自主上报、统计分析能耗等功能。
在另外的一个实施例中，所述检测模块包括光敏传感器、声音传感器和红外传感器。
在另外的一个实施例中，所述主处理器是STM32F030微处理器。
传感器模块(检测模块)，主要由光敏传感器、声音传感器和红外传感器三部分组成。三者综合进行判断，可有效检测到人和车辆的经过，同时减少干扰信号对系统的影响。
微处理器：本系统在智能控制模块中使用的主处理器是STM32F030微处理器。功耗低、性能高，实用性强，性价比较高等优点，适合应用。
软件设计：
LoRa无线通信模块：在本系统中，LoRa模块与GPRS模块和单片机相连，把从web端接收到的指令发送到STM32控制器，实现对路灯的远程控制，同时可以将路等状态数据传送至服务器，实现对路灯状态的监视和分析。
GPRS模块：在本系统中，由于智能路灯的覆盖范围十分广泛，并且需要与控制终端实现通信，采用GPRS通信技术，增加了网络的可靠性与数据传输方便。
控制与监控系统模块：本系统包含前端web系统与后端信息处理系统。前端web系统以网页的形式对系统管理员进行展示，用以控制与系统相关联的智慧路灯以及对路灯状态的监控。后端信息处理系统对远端硬件发来的信息与前 端网页发出的信息进行处理、发送、保存等处理。
在另外的一个实施例中，所述主处理器根据检测模块检测到的数据控制路灯的算法是带补偿的模糊神经网络算法。
在另外的一个实施例中，所述带补偿的模糊神经网络算法包括输入层、模糊化层、补偿模糊推理层和反模糊化层；
所述输入层的各个节点直接与输入向量相连接，输入量包括了时间、地点、光照强度、行人、车辆等；所述模糊化层的每一个节点代表一个语言变量值，其作用是计算各输入向量属于各语言变量值模糊集合的隶属函数；所述补偿模糊推理层的每一个节点代表一条模糊规则，其作用是匹配模糊规则、计算出每条规则的适用度，并进行补偿运算；所述反模糊化层给出输出量开灯信息。
层与层之间是依据模糊逻辑系统的语言变量、补偿模糊推理方法、以及反模糊化函数所构建，其结构如图1所示。
网络的参数确定方法：
(1)隶属度函数的参数确定
输入空间x的论域为
Figure PCTCN2018092617-appb-000001
输出空间y的论域为
Figure PCTCN2018092617-appb-000002
那么对输入空间x的模糊分割可按以下的原则进行：在y随x变化而变化剧烈的子空间对x细分，在y随变化而变化平缓的子空间对x粗分。具体来说，可分析y随x的变化情况，用x在某子空间变化时y的极值点和拐点数(两个端点既不作极值点，也不作拐点)多少来确定y在该子空间的变化剧烈程度，极值点和拐点数越多的子空间，对x的模糊分割越多，反之亦然。首先根据样本确定x和y的论域内近似极值点和拐点数N，及其对应的输入向量
Figure PCTCN2018092617-appb-000003
(其中i表示输入层节点i＝1，…n，k表示极值和拐点对应的模糊分割点k＝1，…，N+2)的值。然后将该输入向量
Figure PCTCN2018092617-appb-000004
作为每维输入变量的模糊分割点，那么每维输入变量的模糊分割数为N+2。
由于模糊分割点可能不是输入论域中的均分点，采用高斯、铃形函数作隶属函数，对网络中各参数初值的选择非常苛刻，否则网络训练会陷入混乱状态，达不到输出误差精度要求，故这里选用三角函数作为隶属函数。
Figure PCTCN2018092617-appb-000005
其中，a ij，b ij，c ij的选择应满足
Figure PCTCN2018092617-appb-000006
隶属函数中的参数值确定后，不再进行修正。
(2)模糊规则数的确定
设第二层中，各输入变量对应的模糊分割的节点数已确定，分别为m i(i＝1，2，…n)那么网络模型的规则数，即第三层节点数为：M＝m 1×m 2×…×m n。显然，若输入变量多，且各输入变量的模糊分割较多，则M是一个很大的数，这必定延长模型的计算速度，由此可知，若能减小M的值，就可大大提高模型的计算速度。
通过分析可知，在逼近每一个样本的计算中，对模型每一个输入变量，其对应的模糊语言变量值的隶属度函数
Figure PCTCN2018092617-appb-000007
中最多有两个不为零，规则层即第三层的输出最多有2 i(i＝1，2，…n)个不为零，这表示最多有2 i(i＝1，2，…n)条规则有效。在计算模型的输出时，可只保留这几个节点，略去其他节点，或者说其他节点不参与计算。为了计算这几个节点的位置，首先作如下假设，设在逼近每一个样本的计算中，网络输入为x i(i＝1，2，…n)，第二层的输出为
Figure PCTCN2018092617-appb-000008
那么有
若
Figure PCTCN2018092617-appb-000009
则以其为输入端的第三层中相应节点的输出值为零，这时可不计算第三层中该节点值，即略去该节点。
若
Figure PCTCN2018092617-appb-000010
则以其为输入端的第三层中相应节点的输出值不为零，此时需计算第三层中该节点值。第三层这些节点的位置大小可由下式来表示：
Figure PCTCN2018092617-appb-000011
式中
p (3)—第三层节点输出值不为零的节点位置大小；
m i—第一层各节点的模糊分割数；
Figure PCTCN2018092617-appb-000012
—第二层节点输出值不为零的节点位置(由隶属函数选取规则知，第一层每个节点的隶属函数值最多有两个不为零，则i＝1，2，…n，j＝1，2)。
按上面的方法确定网络结构后，在逼近每一个样本的计算中，规则总数最多为M'＝2 i(i＝1,2,…n)。所以发明中提出的第三层中的节点要比传统结构要少得多，从而减少了计算量，提高了计算速度。
(3)最后一层的参数值确定
本发明网络本质上也是一种多层前馈网络，故可以仿照前馈网络用误差反传的方法来设计调整参数的学习算法。
设取误差代价函数为：
Figure PCTCN2018092617-appb-000013
式中：
y di——网络的期望输出；
y i——网络的实际输出；
E i——误差函数。
下面给出用误差反传算法来计算
Figure PCTCN2018092617-appb-000014
然后利用一阶梯度寻优算法来调节
Figure PCTCN2018092617-appb-000015
在求得所需的一阶梯度后，最后给出参数调整的学习算法为
Figure PCTCN2018092617-appb-000016
式中β——学习率(0<β<1)。
学习算法：
具有n个输入、1个输出的补偿模糊逻辑系统的m条if-then规则，可表示如下：
FR (k):if x 1is
Figure PCTCN2018092617-appb-000017
and…and x n is
Figure PCTCN2018092617-appb-000018
then y is B k
(i＝1,2,…n k＝1,2,…m)
其中
Figure PCTCN2018092617-appb-000019
是论域U上的模糊集；B k是论域V上的模糊集；x i和y是语言变量；。
对于论域U中一个输入模糊子集A′，根据第k个模糊规则，能够在输出论域V中产生一个输出模糊子集B′。模糊推理采用代数积(·)运算，则由模糊推理规则所导出的V上的模糊集合B’为
Figure PCTCN2018092617-appb-000020
模糊蕴涵采用积运算R p＝A→B
即μ A→B(x,y)＝μ A(x)μ B(y)
定义补偿运算为
Figure PCTCN2018092617-appb-000021
其中n为输入向量的维数，γ为补偿度，γ∈[0,1]。
Figure PCTCN2018092617-appb-000022
采用单值模糊化μ' A(x)＝1，则
Figure PCTCN2018092617-appb-000023
由此可定义反模糊化函数为
Figure PCTCN2018092617-appb-000024
因此，该系统由三角形隶属函数、补偿模糊推理以及改进型重心反模糊化器构成。
目标函数为：
Figure PCTCN2018092617-appb-000025
采用动态调整步长的梯度下降法训练系统的输入、输出隶属函数的参数以 及补偿运算的补偿度。其相应的迭代公式为：
Figure PCTCN2018092617-appb-000026
Figure PCTCN2018092617-appb-000027
Figure PCTCN2018092617-appb-000028
由于补偿度γ∈[0,1]，不妨设
Figure PCTCN2018092617-appb-000029
Figure PCTCN2018092617-appb-000030
所以(t为训练步数)
Figure PCTCN2018092617-appb-000031
Figure PCTCN2018092617-appb-000032
Figure PCTCN2018092617-appb-000033
至此，整个网络的构建和训练学习就完成了，可以用于路灯系统的控制。具体智能路灯系统的控制方案如下：
Web端首先选择系统工作模式，包括自动控制模式和实时控制模式，随后通过云服务器经GPRS模块发送指令至LoRa模块，LoRa模块把指令发送到各个路灯节点上的STM32控制器，控制器分析指令并对路灯的状态进行控制。对于两种工作模式，如果选择实时控制模式，则控制器完全按照服务器指令控制 路灯节点，实现服务器对路灯节点的实时控制。如果选择自动控制模式，控制器按照预先设定好的控制策略控制路灯节点，在白天光照度不强的情况下开启路灯，在夜晚凌晨之前路灯保持开启状态，凌晨以后且有人、车通过，则开启路灯，否则关闭。具体开启路灯的条件根据数据融合算法来确定。
总体设计：系统采用低功耗的LoRa技术，把STM32作为核心数据处理芯片，采用光敏、声音、红外和电流传感器，检测环境的光照亮度变化以及行人、车辆的活动状况。STM32芯片根据采集传感器的信息，实现自动调节路灯亮度；通过LoRa进行无线数据传输，将数据传输到系统终端，经数据处理将所测数据在网页显示，Web界面提供查看路灯状态、路灯开关、远程调节亮度、统计分析能耗、故障监控的可视化操作。
基于LoRa技术的智能路灯系统将会大幅减少电力资源，提高照明管理效率，降低维护成本，同时提供便民信息与服务，对建设智慧城市具有重大意义。随着国内广域物联网喷发式的发展和CLAA组织对LoRa应用的积极推动，国内基于LoRa应用的试点将会越来越多地被部署在各行各业，提供优质高效的物联网服务。当前LoRa技术在智能路灯领域的应用基本上处于空白地位，有着广阔的市场前景。
本发明将物联网技术与路灯照明相结合，针对路灯智能化的需求背景，设计了一种基于LoRa技术的多功能LED智能路灯系统，在路灯传统的照明作用基础上，嵌入了传感、无线传输等智能控制模块，使路灯具有自动调节亮度、远程调节亮度、故障自主上报、统计分析能耗等功能。
该发明主要解决传统路灯照明中功能单一化、缺乏智能化的问题，传统路灯照明易受天气状况影响、控制方式缺乏灵活性、能源浪费大、人工巡检耗费人力与时间、不能实时监控运行状况和收集运行数据，因此需要研究多功能的智能路灯系统，采用新型神经网络融合技术，以弥补传统路灯单一化功能的不足。
以上所述实施例的各技术特征可以进行任意的组合，为使描述简洁，未对上述实施例中的各个技术特征所有可能的组合都进行描述，然而，只要这些技术特征的组合不存在矛盾，都应当认为是本说明书记载的范围。
以上所述实施例仅表达了本发明的几种实施方式，其描述较为具体和详细，但并不能因此而理解为对发明专利范围的限制。应当指出的是，对于本领域的普通技术人员来说，在不脱离本发明构思的前提下，还可以做出若干变形和改进，这些都属于本发明的保护范围。因此，本发明专利的保护范围应以所附权利要求为准。

Claims (5)

  1. 一种基于LoRa技术的多功能LED智能路灯系统，其特征在于，包括：
    所述检测模块，所述检测模块检测环境的光照亮度变化以及行人、车辆的活动状况；
    主处理器，所述主处理器根据检测模块检测到的数据控制路灯；
    LoRa无线通信模块，所述LoRa无线通信模块与GPRS模块和主处理器相连，把从web端接收到的指令发送到主处理器，实现对路灯的远程控制，同时检测模块检测到的数据传送至服务器；以及
    系统终端，所述系统终端接收通过LoRa无线通信模块传输的测模块检测到的数据，经数据处理将所测数据在网页显示，Web界面提供查看路灯状态、路灯开关、远程调节亮度、统计分析能耗和故障监控的可视化操作。
  2. 根据权利要求1所述的基于LoRa技术的多功能LED智能路灯系统，其特征在于，所述检测模块包括光敏传感器、声音传感器和红外传感器。
  3. 根据权利要求1所述的基于LoRa技术的多功能LED智能路灯系统，其特征在于，所述主处理器是STM32F030微处理器。
  4. 根据权利要求1所述的基于LoRa技术的多功能LED智能路灯系统，其特征在于，所述主处理器根据检测模块检测到的数据控制路灯的算法是带补偿的模糊神经网络算法。
  5. 根据权利要求4所述的基于LoRa技术的多功能LED智能路灯系统，其特征在于，所述带补偿的模糊神经网络算法包括输入层、模糊化层、补偿模糊推理层和反模糊化层；
    所述输入层的各个节点直接与输入向量相连接，输入量包括了时间、地点、光照强度、行人、车辆等；所述模糊化层的每一个节点代表一个语言变量值，其作用是计算各输入向量属于各语言变量值模糊集合的隶属函数；所述补偿模糊推理层的每一个节点代表一条模糊规则，其作用是匹配模糊规则、计算出每条规则的适用度，并进行补偿运算；所述反模糊化层给出输出量开灯信息。
PCT/CN2018/092617 2018-05-08 2018-06-25 基于LoRa技术的多功能LED智能路灯系统 WO2019214016A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201810432316.6 2018-05-08
CN201810432316.6A CN108419339B (zh) 2018-05-08 2018-05-08 基于LoRa技术的多功能LED智能路灯系统

Publications (1)

Publication Number Publication Date
WO2019214016A1 true WO2019214016A1 (zh) 2019-11-14

Family

ID=63138114

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/092617 WO2019214016A1 (zh) 2018-05-08 2018-06-25 基于LoRa技术的多功能LED智能路灯系统

Country Status (2)

Country Link
CN (1) CN108419339B (zh)
WO (1) WO2019214016A1 (zh)

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110708798A (zh) * 2019-11-15 2020-01-17 哈尔滨理工大学 一种基于模块识别的智能台灯系统
CN112153785A (zh) * 2020-08-20 2020-12-29 安徽极光照明工程有限公司 一种基于环境自适应式景观灯群控系统
CN115209585A (zh) * 2022-07-07 2022-10-18 达那喜(无锡)绿色能源科技有限公司 一种基于5g技术的路灯系统控制方法
CN115480484A (zh) * 2022-09-14 2022-12-16 中国铁塔股份有限公司重庆市分公司 一种面向智慧灯杆的多源信号集成控制方法及装置
CN116847521A (zh) * 2023-06-15 2023-10-03 深圳市雨星科技有限公司 一种智能太阳能路灯控制方法及系统
CN117596755A (zh) * 2023-12-15 2024-02-23 广东瑞峰光电科技有限公司 一种物联网路灯智能控制方法及系统

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109595535B (zh) * 2018-10-17 2020-07-24 宁波索拉彼工贸有限公司 一种新型节能led灯
CN110415653B (zh) * 2019-07-18 2022-01-18 昆山龙腾光电股份有限公司 背光亮度调节系统及调节方法和液晶显示装置
CN111526649A (zh) * 2020-06-01 2020-08-11 沧州师范学院 一种新能源路灯的智慧照明控制方法
CN118283869A (zh) * 2024-05-06 2024-07-02 重庆市市政设计研究院有限公司 智慧城市路灯控制系统

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170127622A1 (en) * 2015-11-10 2017-05-11 Xu Hong Smart control/iot system for agriculture environment control
CN106686867A (zh) * 2017-03-01 2017-05-17 湖南原创智能科技有限公司 基于LoRa技术的智能路灯管理系统
CN106879135A (zh) * 2017-01-16 2017-06-20 上海博昂电气有限公司 一种LoRa无线智能路灯照明控制系统
CN106912154A (zh) * 2017-05-09 2017-06-30 上海商研机电设备有限公司 一种智能路灯控制系统
CN107169993A (zh) * 2017-05-12 2017-09-15 甘肃政法学院 利用公安视频监控模糊图像对目标物进行检测识别方法

Family Cites Families (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102122132A (zh) * 2010-01-11 2011-07-13 北京航空航天大学 一种基于模糊神经网络的用于环境模拟系统的智能控制系统
CN105960051A (zh) * 2016-05-24 2016-09-21 吉林蓝锐电子科技有限公司 一种无级调光led路灯及控制方法
CN106211518B (zh) * 2016-08-26 2018-06-26 特斯联（北京）科技有限公司 一种智能城市照明系统
CN206024182U (zh) * 2016-09-13 2017-03-15 厦门市致创能源技术有限公司 一种路灯控制系统
CN106507536A (zh) * 2016-10-31 2017-03-15 南昌航空大学 一种依托环境数据分时段模糊控制路灯的照明系统和方法
CN206196092U (zh) * 2016-11-28 2017-05-24 晶志力（苏州）智能科技有限公司 远程路灯控制系统
CN107018607A (zh) * 2016-11-28 2017-08-04 贵州大学 一种路灯照明节能控制系统及其控制方法
CN107529250B (zh) * 2017-07-17 2019-02-15 电子科技大学 一种led路灯调控装置及调控方法
CN107438313A (zh) * 2017-08-08 2017-12-05 方晨 大规模智能灯具的监控系统及监控方法
CN207201054U (zh) * 2017-09-26 2018-04-06 湘潭大学 一种路灯亮度控制系统
CN107995761A (zh) * 2017-12-25 2018-05-04 南京邮电大学 一种基于LoRa无线通信技术的智慧路灯管理系统及方法

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170127622A1 (en) * 2015-11-10 2017-05-11 Xu Hong Smart control/iot system for agriculture environment control
CN106879135A (zh) * 2017-01-16 2017-06-20 上海博昂电气有限公司 一种LoRa无线智能路灯照明控制系统
CN106686867A (zh) * 2017-03-01 2017-05-17 湖南原创智能科技有限公司 基于LoRa技术的智能路灯管理系统
CN106912154A (zh) * 2017-05-09 2017-06-30 上海商研机电设备有限公司 一种智能路灯控制系统
CN107169993A (zh) * 2017-05-12 2017-09-15 甘肃政法学院 利用公安视频监控模糊图像对目标物进行检测识别方法

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110708798A (zh) * 2019-11-15 2020-01-17 哈尔滨理工大学 一种基于模块识别的智能台灯系统
CN112153785A (zh) * 2020-08-20 2020-12-29 安徽极光照明工程有限公司 一种基于环境自适应式景观灯群控系统
CN115209585A (zh) * 2022-07-07 2022-10-18 达那喜(无锡)绿色能源科技有限公司 一种基于5g技术的路灯系统控制方法
CN115480484A (zh) * 2022-09-14 2022-12-16 中国铁塔股份有限公司重庆市分公司 一种面向智慧灯杆的多源信号集成控制方法及装置
CN116847521A (zh) * 2023-06-15 2023-10-03 深圳市雨星科技有限公司 一种智能太阳能路灯控制方法及系统
CN116847521B (zh) * 2023-06-15 2024-02-27 深圳市雨星科技有限公司 一种智能太阳能路灯控制方法及系统
CN117596755A (zh) * 2023-12-15 2024-02-23 广东瑞峰光电科技有限公司 一种物联网路灯智能控制方法及系统
CN117596755B (zh) * 2023-12-15 2024-04-16 广东瑞峰光电科技有限公司 一种物联网路灯智能控制方法及系统

Also Published As

Publication number Publication date
CN108419339B (zh) 2020-07-28
CN108419339A (zh) 2018-08-17

Similar Documents

Publication Publication Date Title
WO2019214016A1 (zh) 基于LoRa技术的多功能LED智能路灯系统
Li et al. Intelligent vehicle network system and smart city management based on genetic algorithms and image perception
CN111739284B (zh) 基于遗传算法优化模糊控制的交通信号灯智能配时方法
CN111263497A (zh) 一种基于无线Mesh自组网的智能光配置系统及方法
CN113096418A (zh) 基于边缘计算的交通网红绿灯控制方法、系统及计算机可读存储介质
CN108966448A (zh) 一种基于自适应模糊决策树的灯光动态调控方法
He et al. Illumination control of intelligent street lamps based on fuzzy decision
CN103547033A (zh) 基于无线传感器执行器网的大型智能灯光节能控制方法
Sikdar et al. An energy efficient street lighting framework: ANN-based approach
Shobana et al. IoT based on Smart Traffic Lights and Streetlight System
CN113505346B (zh) 一种基于人工智能的城市路灯数据处理及联合调控系统
CN108001700B (zh) 一种基于无线传感器网络的机场跑道入侵防御系统
CN109890104A (zh) 基于云服务的具备学习能力的太阳能路灯调光控制系统
CN114390762A (zh) 一种基于边缘计算的自适应调光系统及其工作方法
Yu et al. WSN nodes deployment based on artificial fish school algorithm for traffic monitoring system
CN111683429B (zh) 一种基于物联网的综合管廊智能照明控制系统
Tang et al. An Intelligent Fault Diagnosis Method for Street Lamps
Wu et al. Research on a fuzzy algorithm for intelligent power-saving control of urban street lamps
Yin et al. Energy consumption parameter detection of green energy saving building based on artificial fish swarm algorithm
Chen Design of intelligent lighting energy saving system for Expressway Tunnel Based on Internet of things
Hao et al. Intelligent lighting system of urban road based on internet of things
US11892887B2 (en) Methods, systems, and media for traffic management in smart city based on internet of things
CN118382168B (zh) 分布式部署的可调节式led路灯智能照明控制系统及方法
CN115996503B (zh) 一种自优化楼宇照明传感器节能控制系统
CN113822473B (zh) 一种基于多维数据的牵引力能耗合理区间预测方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 18917989

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 18917989

Country of ref document: EP

Kind code of ref document: A1