WO2021007755A1 - 控制方法、装置、系统和计算机可读介质 - Google Patents

控制方法、装置、系统和计算机可读介质 Download PDF

Info

Publication number
WO2021007755A1
WO2021007755A1 PCT/CN2019/096041 CN2019096041W WO2021007755A1 WO 2021007755 A1 WO2021007755 A1 WO 2021007755A1 CN 2019096041 W CN2019096041 W CN 2019096041W WO 2021007755 A1 WO2021007755 A1 WO 2021007755A1
Authority
WO
WIPO (PCT)
Prior art keywords
microservice
control
inference
logic
sensor data
Prior art date
Application number
PCT/CN2019/096041
Other languages
English (en)
French (fr)
Inventor
孙芃
Original Assignee
西门子（中国）有限公司
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 西门子（中国）有限公司 filed Critical 西门子（中国）有限公司
Priority to CN201980096053.9A priority Critical patent/CN113785270A/zh
Priority to PCT/CN2019/096041 priority patent/WO2021007755A1/zh
Publication of WO2021007755A1 publication Critical patent/WO2021007755A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B19/00—Programme-control systems
    • G05B19/02—Programme-control systems electric
    • G05B19/418—Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS] or computer integrated manufacturing [CIM]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/22—Microcontrol or microprogram arrangements
    • G06F9/24—Loading of the microprogram
    • 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
    • Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]

Definitions

  • the present invention relates to the technical field of industrial control, in particular to a control method, device, system and computer readable medium.
  • the embodiments of the present invention provide a control method, device, system, and computer readable medium, which attach an intelligent control device to a traditional device, with little modification to the traditional device, and can flexibly realize the intelligent control function.
  • the first aspect provides a control system for controlling a device.
  • the system may include: a cloud platform configured to determine a control application to the device that needs to be implemented, and determine at least one microservice required to implement the control application, wherein one microservice is used to implement An artificial intelligence AI inference; determining that the at least one microservice implements the logic of the control application.
  • the system also includes: a control device configured to obtain the at least one microservice and the logic from the cloud platform; load the at least one microservice; run the at least one microservice Each item to obtain the sensor data required for the operation of the microservice and execute the AI inference of the microservice based on the acquired sensor data; perform the AI inference result based on each of the at least one microservice, and According to the logic, a control signal for controlling the device is generated.
  • a control device configured to obtain the at least one microservice and the logic from the cloud platform; load the at least one microservice; run the at least one microservice Each item to obtain the sensor data required for the operation of the microservice and execute the AI inference of the microservice based on the acquired sensor data; perform the AI inference result based on each of the at least one microservice, and According to the logic, a control signal for controlling the device is generated.
  • the cloud platform can determine the control application for the device according to user requirements, and determine the microservices required to implement the control application.
  • the control device obtains the logic of the control application and the microservices required to realize the control application from the cloud platform.
  • the AI chip of the control device loads the microservice, collects sensor data when running the microservice, and performs AI inference based on this; and the processor of the control device generates the control signal of the device based on the result of the AI inference according to the logic of the control application. Because the control application can be flexibly configured, there is no need to load unnecessary microservices, and new microservices can be easily loaded, realizing flexible control of the device.
  • the cloud platform is further configured to: for the first microservice in the at least one microservice, perform training based on historical data of sensor data required when the first microservice performs AI inference , To update the first microservice; the control device is further configured to: obtain the updated first microservice from the cloud platform, and load the updated first microservice.
  • the microservices can be configured conveniently and flexibly, so that the results of AI inference are more accurate.
  • control device is further configured to send the control signal to the device; or the control device is further configured to send the control signal to the cloud platform, and the cloud platform is also It is configured to send the control signal to a control server for controlling the device.
  • a suitable control method can be determined according to the existing network architecture and the settings of the equipment to be controlled.
  • a control device for controlling a device, including: at least one sensor configured to collect sensor data; a communication module configured to obtain at least one microservice from a cloud platform; The logic of a control application implemented by the at least one microservice, wherein one microservice is used to implement an artificial intelligence AI inference; an artificial intelligence AI chip is configured to load the at least one microservice Service, run each of the at least one microservice to obtain sensor data required for the operation of the microservice and execute AI inference of the microservice based on the obtained sensor data; a processor configured to be based on Each of the at least one microservice executes the result of AI inference, and generates a control signal for controlling the device according to the logic.
  • the cloud platform can determine the control application for the device according to user requirements, and determine the microservices required to implement the control application.
  • the control device obtains the logic of the control application and the microservices required to realize the control application from the cloud platform.
  • the AI chip of the control device loads the microservice, collects sensor data when running the microservice, and performs AI inference based on this; and the processor of the control device generates the control signal of the device based on the result of the AI inference according to the logic of the control application. Since the control application can be flexibly configured, there is no need to load unnecessary microservices, and new microservices can be easily loaded, realizing flexible control of the device.
  • the communication module is further configured to obtain an updated first microservice, wherein the first microservice is one of the at least one microservice, and the updated first microservice is The microservice is obtained by training based on the historical data of the sensor data required when the first microservice executes AI inference; the AI chip is also configured to load the updated first microservice.
  • the communication module is further configured to obtain an updated first microservice, wherein the first microservice is one of the at least one microservice, and the updated first microservice is The microservice is obtained by training based on the historical data of the sensor data required when the first microservice executes AI inference; the AI chip is also configured to load the updated first microservice.
  • a control method for controlling a device, including: acquiring at least one microservice and a logic of a control application implemented by the at least one microservice, wherein one of the microservices is used for Implement a kind of artificial intelligence AI inference; load the at least one microservice; run each of the at least one microservice to obtain sensor data required for the operation of the microservice and execute AI based on the obtained sensor data Inference; the result of performing AI inference based on each of the at least one microservice, and generating a control signal for controlling the device according to the logic.
  • control application and microservices can be flexibly configured, there is no need to load unnecessary microservices, and new microservices can be easily loaded, thus realizing flexible control of devices.
  • the method further includes: obtaining an updated first microservice, wherein the first microservice is one of the at least one microservice, and the updated first microservice is based on The first microservice is obtained by training the historical data of sensor data required when performing AI inference; and the updated first microservice is loaded.
  • the first microservice is obtained by training the historical data of sensor data required when performing AI inference; and the updated first microservice is loaded.
  • a computer-readable medium stores computer-readable code
  • the computer-readable code is used to control a device, and includes: code of at least one microservice, wherein A microservice is used to implement an artificial intelligence AI inference.
  • the AI chip runs each of the at least one microservice, In order to obtain the sensor data required for the operation of the microservice and perform AI inference based on the obtained sensor data; a logic code that controls the application implemented by the at least one microservice, when the logic code is executed by a processor At the time, based on the result of performing AI inference on each of the at least one microservice, and generate a control signal for controlling the device according to the logic.
  • control application and microservices can be flexibly configured, there is no need to load unnecessary microservices, and new microservices can be easily loaded, thus realizing flexible control of devices.
  • a control device for controlling a device, including: at least one sensor configured to collect sensor data; a memory configured to store codes for a plurality of microservices, and at least one control Each type of code that controls application logic in the application, where one microservice is used to implement an artificial intelligence AI inference, and each type of control should require some or all of the plurality of microservices; a control switch, Is configured to set a control application that needs to be performed on the device; a processor is configured to run the code of the control application logic set by the control switch; an artificial intelligence AI chip is configured to: At least one microservice required to implement the control application set by the control switch is loaded in the memory; each of the at least one microservice is run to obtain sensor data required for the operation of the microservice and based on the acquisition Perform AI inference of the microservice based on the sensor data of the microservice; the processor is also configured to perform AI inference results based on each of the at least one microservice, and generate the results for controlling the microservice according to the logic
  • the operator can set the control switch to configure the control application realized by the control device.
  • the logic of various control applications and the realization of the microservices required to implement the control applications are stored in the memory of the control device.
  • the required logic and microservices are obtained from the memory according to the configuration of the control switch, and the AI inferred by running the microservices As a result, it runs logic to generate control signals to achieve flexible control of the equipment.
  • control device further includes a communication module configured to: send the control signal to the device; or send the control signal to a cloud platform, and the cloud platform is also configured to The control signal is sent to a control server for controlling the device.
  • a suitable control method can be determined according to the existing network architecture and the settings of the equipment to be controlled.
  • Figure 1A and Figure 1B are schematic diagrams of two smart street lights.
  • Figure 2 is a schematic structural diagram of an intelligent control system provided by an embodiment of the present invention.
  • Fig. 3 is a flowchart of an intelligent control method provided by an embodiment of the present invention.
  • Fig. 4 is a schematic structural diagram of an intelligent control device provided by an embodiment of the present invention.
  • Fig. 5 is another schematic structural diagram of the intelligent control device provided by the embodiment of the present invention.
  • Fig. 6 is a schematic diagram of a smart street light control device provided by an embodiment of the present invention installed on a traditional street light.
  • FIG. 7 is a schematic diagram of various microservices loaded on the AI chip in the smart street light control device provided by an embodiment of the present invention.
  • Fig. 8 shows various control applications of street lamp light brightness realized by the intelligent street lamp control device provided by the embodiment of the present invention.
  • Micro base station 1001 Micro base station 1002: Hot spot coverage 1003: Environmental monitoring
  • Control system 10 Control device 20: Controlled equipment
  • AI chip 102 processor 103: memory
  • 5013 People density micro service 5014: Traffic flow micro service 5015: Micro weather judgment micro service
  • Street light control signal 7 Street light control signal 5017: Day and night time setting of micro-services
  • FIG. 1A and Figure 1B show two existing smart street lights.
  • the smart street light shown in FIG. 1A can perform various control applications such as micro base station 1001, hotspot coverage 1002, environmental monitoring 1003, security monitoring 1004, advertising operation 1005, outdoor broadcasting 1006, charging management 1007, and multifunctional gateway 1008.
  • the smart street light shown in FIG. 1B can perform various control applications such as WIFI network 1011, environmental monitoring 1012, video monitoring 1013, information release 1014, smart lighting 1015, network base station 1016, city broadcasting 1017, and car charging 1018.
  • the control application settings of the above two types of smart street lights are fixed and lack flexibility.
  • the embodiment of the present invention provides a control device, which has an intelligent control function, realizes the intelligent control of the equipment by adopting an artificial intelligence algorithm, and improves the flexibility of equipment control. It is attached to traditional equipment without replacing traditional equipment, and its control function can be flexibly configured according to user needs. The realization of its artificial intelligence algorithm and control logic can be flexibly changed without requiring major changes to the equipment.
  • the cloud platform may determine the control application for the device according to user requirements, and determine the microservices required to implement the control application.
  • the control device obtains the logic of the control application and the microservices required to realize the control application from the cloud platform.
  • the AI chip of the control device loads the microservice, collects sensor data when running the microservice, and performs AI inference based on this; and the processor of the control device generates the control signal of the device based on the result of the AI inference according to the logic of the control application. Since the control application can be flexibly configured, there is no need to load unnecessary microservices, and new microservices can be easily loaded, realizing flexible control of the device.
  • control device has a control switch, and the operator can set the control switch to configure the control application realized by the control device.
  • the logic of various control applications and the realization of the microservices required to implement the control applications are stored in the memory of the control device.
  • the required logic and microservices are obtained from the memory according to the configuration of the control switch, and the AI inferred by running the microservices As a result, it runs logic to generate control signals to achieve flexible control of the equipment.
  • a street lamp control device is taken as an example to illustrate the loading of microservices, the operation of control application logic, and the generation and transmission of control signals.
  • a specific control application is taken as an example to illustrate the flexibility of the control device realized by the embodiment of the present invention in realizing the intelligent control function.
  • FIG. 2 is a schematic structural diagram of an intelligent control system provided by an embodiment of the present invention. If shown in 2, the control system 100 is used to realize the intelligent control of the device 20, including:
  • a control device 10 is configured to generate control signals for the device 20.
  • the control device 10 may be installed near the equipment 20.
  • the street light control device 50 in FIG. 6 adopts a pole-holding installation method and is installed on the pole of a street light.
  • the control device 10 collects sensor data through its own sensor 104.
  • a temperature sensor is used to collect the temperature near the device 20
  • a humidity sensor is used to collect the humidity near the device 20
  • a camera is used to take pictures of the environment near the device 20.
  • the control device 10 performs AI inference based on the collected sensor data, and then generates a control signal of the device 20.
  • control functions implemented by the control device 10 can be flexibly configured.
  • the control device 10 obtains the logic of the control application required by the user and at least one microservice required to implement the logic from the cloud platform 30 in the control system 100, and the control device 10 loads and runs the microservice to obtain sensor data and proceed accordingly.
  • AI inference The control device 10 generates a control signal of the device 20 after processing according to the acquired logic of the control application based on the result of AI inference. Since the logic and microservices of the control application can be flexibly configured, flexible control of the device 20 is realized.
  • the control signal generated by the control device 10 may be sent to the control server 40 of the device 20 via the cloud platform 30, and then sent to the controller of the device 20 after conversion by the control server 40, so as to realize the control of the device 20.
  • the control signal generated by the control device 10 can be sent to the control server 40, and the control server 40 is transformed and sent to the controller of the device server (30) to realize the control of the device 20.
  • the control signal generated by the control device 10 is directly sent to the controller of the device 20 to realize the control of the device 20. Since different devices have different formats for reading information, it may be necessary to change the format of the control signal during the forwarding process of the control signal.
  • the control signal generated by the control device 10 may be used to control all the devices 20 in an area.
  • control device 10 is deployed on some devices 20 in an area, and the generated control signal can be used for all the devices 20 in the area. Take control.
  • the control server 40 converts the control signal into a signal for each device 20 and transmits it one by one.
  • the control device 10 can communicate with the cloud platform 30 or the controller of the device 20 in a wireless communication manner to realize the transmission of control signals.
  • the available wireless communication methods include but are not limited to WiFi, the fourth generation mobile communication 4G, LoRa, etc.
  • control system 100 shown in FIG. 2 there is no need to replace traditional devices with new devices with intelligent control functions, and the setting of control functions is flexible.
  • the logic of the control application, or the AI inference model or decision threshold, timing relationship, etc. of the microservice is changed, the update of the control application can also be easily realized by loading the updated logic and microservice.
  • Fig. 3 is a flowchart of an intelligent control method provided by an embodiment of the present invention. As shown in Figure 3, the method may include the following steps:
  • the cloud platform 30 in the control system 100 determines a control application to the device 20 that needs to be implemented.
  • the control system 100 can determine the control application according to user requirements. For example, if the user needs to realize the automatic adjustment of the brightness of the street lamp and the monitoring of the traffic flow, the control system 100 determines that the control application includes: a brightness adjustment application and a traffic flow monitoring application.
  • each control application is implemented according to a certain logic and needs to be based on some results of AI inference.
  • AI inference is implemented in a microservice manner.
  • the microservices required by the brightness adjustment application include the visibility calculation microservice and the road condition judgment microservice, which are used to determine the visibility of the road section near the street light and the road condition judgment of the road section near the street light; while the microservice required by the traffic flow monitoring application includes the traffic flow microservice.
  • the service is used to determine the real-time traffic flow of the section near the street light.
  • the following simple example illustrates the logic of the brightness adjustment application.
  • the brightness of the street lamp is determined according to the visibility of the road section near the street lamp and the road condition of the road section near the street lamp.
  • the visibility calculation microservice When the visibility calculation microservice performs AI inference, the result is that the visibility of the road section near the street lamp is less than the first A visibility threshold, and the result of AI inference by the road condition judgment microservice determines that there is micro weather in the road near the street light, then the brightness of the street light is determined to be the first brightness; when the visibility calculation microservice performs AI inference, the result is the visibility of the road near the street light It is not less than the first visibility threshold, and when the AI inference result of the road condition judgment microservice determines that there is no micro weather in the road section near the street light, the brightness of the street light is determined to be the second brightness, where the second brightness is less than the first brightness, ....
  • the control device 10 obtains the logic of microservices and control applications from the cloud platform 30.
  • control device 10 downloads the logic of the control application determined by the cloud platform 30 and the code of each microservice required to implement the logic to the local.
  • the control device 10 can establish a wireless connection with the cloud platform 30 by means of wireless communication and download codes accordingly.
  • the control device 10 runs various microservices. During the operation of each microservice, the sensor data required for the operation of the microservice is acquired, and the AI inference of the microservice is executed based on the acquired sensor data. For examples of performing AI inference based on sensor data, see Figure 7, Figure 8 and the corresponding description. It should be noted that the microservice can perform AI inference based on the data collected from the sensor during the operation process, and can also perform further AI inference based on the results of other microservices performing AI inference.
  • the control device 10 generates a control signal for controlling the device 20 according to the logic of the above-mentioned control application according to the result of AI inference of each microservice.
  • the control device 10 sends a control signal to the device 20.
  • Step S305 has multiple ways including the following implementation ways:
  • Method 1 The control device 10 sends the control signal directly to the device 20;
  • Manner 2 The control device 10 sends the control signal to the cloud platform 30, and the cloud platform 30 sends the control signal to the control server 40, and the control server 40 sends the control signal to the controller of the device 20, thereby realizing the control of the device 20 .
  • the cloud platform 30 can also use its own computing resources to update various microservices.
  • a microservice can be used to perform AI inference with historical data of sensor data for training to obtain updated mathematical models, decision thresholds, timing relationships, etc., and the control device 10 can periodically or when microservices are updated from the cloud The updated microservice is obtained at the platform 30.
  • the update of microservices can further ensure the accuracy of AI inference results.
  • Fig. 4 is a schematic structural diagram of an intelligent control device provided by an embodiment of the present invention. As shown in FIG. 4, the control device 10 may include:
  • At least one sensor 104 is configured to collect sensor data.
  • the sensor 104 can be located inside the control device 10 in terms of hardware structure, and can be located outside the control device 10, and transmits sensor data with the control device 10 through wireless communication.
  • the sensor 104 collects data of the device 20, such as temperature, humidity, motor speed, liquid flow speed, etc., and the collected sensor data is used for subsequent AI inference.
  • the communication module 106 is configured to communicate with other devices other than the control device 10, such as: obtaining microservices from the cloud platform 30 and controlling application logic implemented by the obtained microservices.
  • a microservice is used to implement a kind of AI inference, and the cloud platform 30 determines the control application to be executed by the control device 10 according to the user’s needs, as well as the various microservices required for the implementation of the control application, and the control
  • the communication module 106 of the device 10 is responsible for obtaining the logic of the control application to be executed and the codes of various microservices from the cloud platform 30.
  • the control device 10 can execute one type of control application, and can also execute multiple control applications. Which control applications to execute is determined by the cloud platform 30, and the control device 10 only needs to obtain the code of logic and microservices.
  • the artificial intelligence AI chip 101 is configured to load various microservices obtained by the communication module 104 and run various microservices. During the operation of a microservice, the sensor data required for the operation of the microservice is obtained from the sensor 104 and the AI inference of the microservice is executed based on the obtained sensor data.
  • the processor 102 is configured to execute AI inference results based on each of the microservices required for the execution of a control application, and generate a control signal for controlling the device 20 according to the logic of the control application.
  • the AI inference and control application logic are executed by the AI chip and the processor.
  • the advantage is that the AI chip usually has strong computing power, which can meet the computing power requirements of AI inference, and has good real-time performance.
  • the processor can be used for logic. Control and interface with various peripheral devices, such as communication modules.
  • the communication module 104 is also configured to obtain updated microservices from the cloud platform 30 under the control of the processor 102.
  • the cloud platform 30 performs training based on historical data of sensor data required by a microservice when performing AI inference, so as to update the data model, decision threshold, timing relationship, etc. used in AI inference.
  • the AI chip 101 is also configured to load the updated first microservice and run the updated first microservice to perform AI inference.
  • the communication module 104 may periodically obtain updated microservices from the cloud platform 30, or may obtain updated microservices after receiving an update notification from the cloud platform 30.
  • Fig. 5 is another schematic structural diagram of the intelligent control device provided by the embodiment of the present invention.
  • the control device 10 shown in FIG. 5 also includes a control switch 105 configured to set control applications that need to be executed on the device 20, such as brightness adjustment applications, traffic flow monitoring applications, and the like. All the logic of the control application and the code of each microservice required for the implementation of the control application are pre-stored in the memory 103 of the control device 10, or obtained in advance from the cloud platform 30 through the communication module 10 and stored in the memory 103 in. The operator sets the control application that needs to be executed by the control device 10 by setting the control switch 105.
  • the processor 102 reads the value set by the control switch 105, determines the control application that needs to be executed by the control device 10, and stores the data from the memory 103.
  • the code of a certain control application is found in the code of the control application logic, and the AI chip 101 finds the code of the microservice required by the control application from the codes of multiple microservices stored in the memory 103, loads and runs these microservices. service.
  • Fig. 6 is a schematic diagram of a smart street light control device provided by an embodiment of the present invention installed on a traditional street light.
  • the street light control device 50 is installed on the light pole of a traditional street light by holding a pole, and includes a humidity sensor 5041, a temperature sensor 5042, a camera 5043 and a radar 5044.
  • each sensor can adopt a plug-and-play mode, which can be flexibly configured according to user needs.
  • the processor 502 of the street light control device 50 obtains the logic code of the control application and the code 5 of the micro service required to realize the control application from the cloud platform 80.
  • the AI chip 501 of the street light control device 50 records the code of the micro service, and runs the micro service. Service to read the data of each sensor and perform AI inference The result 1 of the AI inference is sent to the processor 502, and the processor 502 performs processing in accordance with the previously acquired logic to obtain the street light control signal 2 or the street light control signal 7. Among them, the street light control signal 2 is sent to the cloud platform 80, and then processed by the cloud platform 80 to form the street light control signal 3, and sent to the street light control server 70, such as the street light control server used by all street lights in the road section where the street light is located. 70.
  • the street lamp control server 70 converts the street lamp control signal 3 into a street lamp control signal 4 for controlling the street lamp, and sends it to the controller 60 of the street lamp. Another implementation manner is that the street light control signal 7 is directly sent by the street light control device 50 to the street light controller 60 for controlling the street light.
  • the sensor can be plug-and-play according to user needs, and on the software, various microservices can be flexibly loaded and run and disabled.
  • FIG. 7 is a schematic diagram of various microservices loaded on the AI chip in the smart street light control device provided by an embodiment of the present invention. As shown in Figure 7, the following microservices are loaded on the AI chip 101:
  • Visibility calculation microservice 5011 used to calculate the visibility of the air near street lights based on the pictures or videos collected by the camera 5043;
  • the road condition judgment microservice 5012 is used to perform AI inference based on the pictures or videos collected by the camera 5043 and determine the road conditions near the street lights, such as: ice and snow roads, obstacles on the roads, etc.;
  • Pedestrian density microservice 5013 used to calculate the density of pedestrians near street lights based on the pictures or videos collected by the camera 5043;
  • the traffic flow microservice 5014 is used to calculate the traffic flow near the street light according to the pictures or videos collected by the camera 5043, or calculate the traffic flow near the street lights according to the results of the radar 5044 scanning, or according to the pictures or videos collected by the camera 5043 and the radar 5044 The result of scanning calculates the traffic flow near the street light;
  • the micro-weather judgment microservice 5015 is used to calculate the visibility of the air near the street light calculated by the micro-service 5011 and the road condition judgment micro-service 5012 calculated based on the air humidity near the street lamp collected by the humidity sensor 5041, the temperature near the street lamp collected by the temperature sensor 5042, and the visibility calculation microservice Information about road conditions near roads to determine whether micro-weather conditions have occurred, such as rain, snow, fog, etc.;
  • the special event judgment microservice 5016 is used for the traffic flow near the street light calculated by the traffic flow microservice 5014, the pedestrian density near the street light calculated by the pedestrian density microservice 5013, and the pedestrian density and traffic obtained by the cloud platform 80 based on training
  • the normal mode 6 of the flow is to determine whether a special event has occurred near the street light, such as long-term traffic congestion (the duration of traffic congestion exceeds the preset duration threshold), etc.
  • the cloud platform 80 can be trained to obtain the normal mode 6 based on the historical value of the data collected by the radar 5044 and/or the camera 5043.
  • the cloud platform 80 may also train to obtain the normal mode 6 based on the historical value of the traffic flow near the street light calculated by the traffic flow microservice 5014 and the historical value of the pedestrian density near the street light calculated by the pedestrian density microservice 5013.
  • the result 1 of AI inference performed by the micro-weather judgment microservice 5015 and the special event judgment microservice 5016 is sent to the processor 502 for generating the street light control signal 2.
  • FIG. 8 shows examples of several microservices that can be loaded by the street light control device provided by the embodiment of the present invention.
  • different microservices can be applied to different scenarios and can control the brightness of street lights according to the corresponding logic.
  • These microservices can be flexibly modified according to user needs, and the corresponding sensors can also be plug-and-played to achieve flexible hardware configuration .
  • the night start and end times are different. These data can be obtained in advance by the cloud platform 80 or the street light control device 50 from a weather agency or website, and stored in the street light control device 50 in advance. If there is a change request (for example, the government wants a street light to turn on before dark hours), the cloud platform 80 can also update the microservice on a yearly or monthly basis.
  • Adjustments can be based on weekly strategies, such as analyzing historical data on pedestrian density of traffic flow to determine the normal pattern of traffic flow and pedestrian density near a street light.
  • the adjustment can also be based on a more flexible mode, such as: for example, when a vehicle or pedestrian is observed under the light of a street light, the brightness of the street light is increased to minimize potential accidents and reduce crime rates.
  • micro-weather judgment microservice 5015 can be loaded, the road near the street lamp can be brighter, and it can also be used as a warning to the car driver, which can greatly improve the road safety level.
  • This microservice can be updated daily.
  • the street light control device 50 can also load a special event judgment microservice 5016. For example, during long holidays each year, traffic jams usually occur towards the end of the holiday. In this special case, street lights may need to be kept longer and brighter. However, during the holidays, certain road sections may be Usually the traffic flow is much less, at this time you can control the street lights to switch to a more energy-saving mode. This microservice can be updated hourly.
  • the middle of Figure 8 shows the above-mentioned microservices that can be flexibly loaded, and the left side correspondingly shows the types of sensors required to run a microservice.
  • the top part represents the server of a weather agency or website, and the day and night time settings
  • the microservice 5017 can obtain the night start and end time of the city where the street lamp is located on different dates from the server during operation.
  • the histogram on the right side of the microservice shown in FIG. 8 shows the street lamp brightness at different time periods of the day calculated based on the result of the corresponding microservice AI inference.
  • an embodiment of the present invention also provides a computer-readable medium that stores computer-readable codes, and the computer-readable codes are used to control a device to be controlled (such as a street lamp).
  • the computer readable code may include:
  • the code of at least one microservice where one microservice is used to implement an artificial intelligence AI inference.
  • the AI chip runs each of the at least one microservice One to obtain the sensor data required for the operation of the microservice and perform AI inference based on the obtained sensor data;
  • a code that controls the logic of an application implemented by at least one microservice When the logic code is executed by a processor, it executes the result of AI inference based on each of the at least one microservice, and generates a control band according to the logic. Control signal for control equipment.
  • Examples of computer-readable media include floppy disks, hard disks, magneto-optical disks, optical disks (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW, DVD+RW), magnetic tape, Volatile memory card and ROM.
  • the computer-readable instructions can be downloaded from a server computer or a cloud via a communication network.
  • An example of a computer-readable medium is a storage device on the cloud platform 30.
  • Another example of the computer-readable medium is the memory 103 of the control device 10.
  • system structure described in the foregoing embodiments may be a physical structure or a logical structure. That is, some modules may be implemented by the same physical entity, or some modules may be implemented by at least two physical entities, or at least Some components in two independent devices are implemented together.
  • the hardware unit can be implemented mechanically or electrically.
  • a hardware unit may include permanent dedicated circuits or logic (such as dedicated processors, Field-Programmable Gate Array (FPGA) or Application Specific Integrated Circuits (ASIC), etc.). Complete the corresponding operation.
  • the hardware unit may also include programmable logic or circuits (such as general-purpose processors or other programmable processors), which may be temporarily set by software to complete corresponding operations.
  • the specific implementation mode mechanical method, or dedicated permanent circuit, or temporarily set circuit

Landscapes

  • Engineering & Computer Science (AREA)
  • General Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Software Systems (AREA)
  • Theoretical Computer Science (AREA)
  • Manufacturing & Machinery (AREA)
  • Quality & Reliability (AREA)
  • Automation & Control Theory (AREA)
  • Circuit Arrangement For Electric Light Sources In General (AREA)

Abstract

一种控制方法、装置、系统和计算机可读介质，用以灵活实现对设备的智能控制功能。一种控制系统(100)包括：一个云平台(30)，被配置为确定需要实现的对设备(20)的一种控制应用、实现所述控制应用所需的至少一项微服务，以及所述控制应用的逻辑；一个控制装置(10)，被配置为从所述云平台(30)处获取并加载运行所述至少一项微服务以及所述逻辑，执行各项微服务的AI推断；基于AI推断的结果按照所述逻辑生成设备(20)的控制信号。

Description

控制方法、装置、系统和计算机可读介质 技术领域
本发明涉及工业控制技术领域，尤其涉及一种控制方法、装置、系统和计算机可读介质。
背景技术
在已建造的基础设施、设备上增加智能控制的功能，提高控制的灵活性已成为一种趋势。比如采用智能路灯来取代传统路灯，使用人工智能解决方案来控制路灯亮度，可节省高达30％以上的能源成本。
一种实现方式是使用具有智能控制功能的设备来替代现有的传统设备，但通常耗资较大，且会影响设备使用，为人们工作和生活带来不便。比如：用智能路灯替代传统路灯通常需要挖出传统路灯并安装智能路灯，而且还可能损坏道路基础设施，影响人们的交通出行。而挖出的传统路灯即使运行良好，由于不再使用，其硬件成本也被浪费。
发明内容
本发明实施例提供了一种控制方法、装置、系统和计算机可读介质，将智能控制设备附加到传统设备上，对传统设备的改动不大，能够灵活地实现智能控制的功能。
第一方面，提供一种控制系统，用于对一个设备进行控制。该系统可包括：一个云平台，被配置为确定需要实现的对所述设备的一种控制应用，确定实现所述控制应用所需的至少一项微服务，其中，一项微服务用于实现一种人工智能AI推断；确定所述至少一项微服务实现所述控制应用的逻辑。该系统还包括：一个控制装置，被配置为从所述云平台处获取所述至少一项微服务以及所述逻辑；加载所述至少一项微服务；运行所述至少一项微服务中的每一项，以获取该项微服务运行所需传感器数据并基于获取的传感器数据执行该项微服务的AI推断；基于所述至少一项微服务中的每一项执行AI推断的结果，并按照所述逻辑生成用于控制所述设备的控制信号。
其中，云平台可根据用户需求确定对设备的控制应用，并确定实现该控制应用所需的微服务。控制装置从云平台处获取该控制应用的逻辑和实现该控制应用所需的微服务。控制装置的AI芯片加载微服务，在运行微服务时采集传感器数据并据此进行AI推断；而控制装置的处理器按照该控制应用的逻辑，基于AI推断的结果生成设备的控制信号。由于控制应用可 灵活配置，无需加载不需要的微服务，且可容易地加载新的微服务，实现了对设备的灵活控制。
可选地，所述云平台，还被配置为：针对所述至少一项微服务中的第一微服务，基于所述第一微服务执行AI推断时所需的传感器数据的历史数据进行训练，以更新所述第一微服务；所述控制装置，还被配置为：从所述云平台处获取更新后的所述第一微服务，并加载更新后的所述第一微服务。这样就实现了微服务的灵活更新和配置，即使控制装置已经投入使用，也可以方便灵活地配置微服务，从而AI推断的结果更准确。
可选地，所述控制装置还被配置为将所述控制信号发送至所述设备；或者所述控制装置还被配置为将所述控制信号发送至所述云平台，且所述云平台还被配置为将所述控制信号发送至用于控制所述设备的控制服务器。其中，可根据已有网络架构以及待控制设备的设置来确定适合的控制方式。
第二方面，提供一种控制装置，用于对一个设备进行控制，包括：至少一个传感器，被配置为采集传感器数据；一个通信模块，被配置为从一个云平台处获取至少一项微服务以及所述至少一项微服务所实现的一种控制应用的逻辑，其中，一项微服务用于实现一种人工智能AI推断；一个人工智能AI芯片，被配置为：加载所述至少一项微服务，运行所述至少一项微服务中的每一项，以获取该项微服务运行所需传感器数据并基于获取的传感器数据执行该项微服务的AI推断；一个处理器，被配置为基于所述至少一项微服务中的每一项执行AI推断的结果，并按照所述逻辑生成用于控制所述设备的控制信号。
其中，云平台可根据用户需求确定对设备的控制应用，并确定实现该控制应用所需的微服务。控制装置从云平台处获取该控制应用的逻辑和实现该控制应用所需的微服务。控制装置的AI芯片加载微服务，在运行微服务时采集传感器数据并据此进行AI推断；而控制装置的处理器按照该控制应用的逻辑，基于AI推断的结果生成设备的控制信号。由于控制应用可灵活配置，无需加载不需要的微服务，且可容易地加载新的微服务，实现了对设备的灵活控制。
可选地，所述通信模块，还被配置为获取更新后的第一微服务，其中所述第一微服务为所述至少一项微服务中的一项，且更新后的所述第一微服务是基于所述第一微服务执行AI推断时所需的传感器数据的历史数据进行训练得到的；所述AI芯片，还被配置为加载所述更新后的所述第一微服务。这样就实现了微服务的灵活更新和配置，即使控制装置已经投入使用，也可以方便灵活地配置微服务，从而AI推断的结果更准确。
第三方面，提供一种控制方法，用于对一个设备进行控制，包括：获取至少一项微服务以及所述至少一项微服务实现的一种控制应用的逻辑，其中一项微服务用于实现一种人工智能AI推断；加载所述至少一项微服务；运行所述至少一项微服务中的每一项，以获取该项微服务运行所需传感器数据并基于获取的传感器数据执行AI推断；基于所述至少一项微服务中的每一项执行AI推断的结果，并按照所述逻辑生成用于控制所述设备的控制信号。
其中，由于控制应用和微服务可灵活配置，无需加载不需要的微服务，且可容易地加载新的微服务，实现了对设备的灵活控制。
可选地，该方法还包括：获取更新后的第一微服务，其中所述第一微服务为所述至少一项微服务中的一项，且更新后的所述第一微服务是基于所述第一微服务执行AI推断时所需的传感器数据的历史数据进行训练得到的；加载更新后的所述第一微服务。这样就实现了微服务的灵活更新和配置，从而AI推断的结果更准确。
第四方面，提供一种计算机可读介质，所述计算机可读介质存储有计算机可读代码，所述计算机可读代码用于对一个设备进行控制，包括：至少一项微服务的代码，其中，一项微服务用于实现一种人工智能AI推断，当所述至少一项微服务被加载到一个AI芯片上时，所述AI芯片运行所述至少一项微服务中的每一项，以获取该项微服务运行所需传感器数据并基于获取的传感器数据执行AI推断；所述至少一项微服务实现的一种控制应用的逻辑的代码，当所述逻辑的代码被一个处理器执行时，基于所述至少一项微服务中的每一项执行AI推断的结果，并按照所述逻辑生成控制所述设备的控制信号。
其中，由于控制应用和微服务可灵活配置，无需加载不需要的微服务，且可容易地加载新的微服务，实现了对设备的灵活控制。
第五方面，提供一种控制装置，用于对一个设备进行控制，包括：至少一个传感器，被配置为采集传感器数据；一个存储器，被配置为存储复数项微服务的代码，以及至少一种控制应用中每一种控制应用逻辑的代码，其中，一项微服务用于实现一种人工智能AI推断，实现每一种控制应需要所述复数项微服务中的部分或全部；一个控制开关，被配置为设置需要对所述设备进行的一种控制应用；一个处理器，被配置为运行所述控制开关所设置的控制应用逻辑的代码；一个人工智能AI芯片，被配置为：从所述存储器中加载实现所述控制开关所设置的控制应用所需的至少一项微服务；运行所述至少一项微服务中的每一项，以获取该项微服务运行所需传感器数据并基于获取的传感器数据执行该项微服务的AI推断；所述处理器， 还被配置为基于所述至少一项微服务中的每一项执行AI推断的结果，并按照所述逻辑生成用于控制所述设备的控制信号。
其中，操作人员可设置该控制开关以配置该控制装置所实现的控制应用。各种控制应用的逻辑和实现控制应用所需的微服务实现存储在控制装置的存储器中，根据该控制开关的配置从存储器中获取所需的逻辑和微服务，通过运行微服务得到AI推断的结果，并运行逻辑生成控制信号，来实现对设备的灵活控制。
可选地，所述控制装置还包括一个通信模块，被配置为：将所述控制信号发送至所述设备；或者将所述控制信号发送至一个云平台，且所述云平台还被配置为将所述控制信号发送至用于控制所述设备的控制服务器。其中，可根据已有网络架构以及待控制设备的设置来确定适合的控制方式。
附图说明
图1A和图1B为两种智能路灯的示意图。
图2为本发明实施例提供的智能控制系统的结构示意图。
图3为本发明实施例提供的智能控制方法的流程图。
图4为本发明实施例提供的智能控制装置的结构示意图。
图5为本发明实施例提供的智能控制装置的又一结构示意图。
图6为本发明实施例提供的智能路灯控制装置安装在传统路灯上的示意图。
图7为本发明实施例提供的智能路灯控制装置中AI芯片上加载的各项微服务的示意图。
图8示出了本发明实施例提供的智能路灯控制装置实现的路灯灯光亮度的各种控制应用。
附图标记列表：
1001：微基站                1002：热点覆盖             1003：环境监测
1004：安防监控              1005：广告运营             1006：户外广播
1007：充电管理              1008：多功能网关           1011：WIFI网络
1012：环境监测              1013：视频监控             1014：信息发布
1015：智能照明              1016：网络基站             1017：城市广播
1018：汽车充电
100：控制系统               10：控制装置               20：被控制的设备
30：云平台                  40：设备20的控制服务器
101：AI芯片                 102：处理器                103：存储器
104：传感器                 105：控制开关              106：通信模块
1011：微服务                1021：控制应用的逻辑
S300：确定控制应用和微服务  S301：获取微服务及控制应用的逻辑 S302：加载微服务
S303：运行微服务            S304：生成控制信号         S305：发送控制信号
50：路灯控制装置            501：AI芯片                502：处理器
5041：湿度传感器            5042：温度传感器           5043：摄像头
5044：雷达                  60：路灯的控制器           70：路灯的控制服务器
80：云平台                  5011：能见度计算微服务     5012：路况判断微服务
5013：人密度微服务          5014：交通流量微服务       5015：微天气判断微服务
5016：特殊事件判断微服务    1：AI推断的结果            6：正常模式
2：路灯控制信号             5：微服务及控制应用的逻辑  3：路灯控制信号
4：路灯控制信号             7：路灯控制信号            5017：昼夜时间设置微服务
具体实施方式
如前所述，使用具有智能控制功能的设备来替代传统设备，具有耗资巨大，使用不便等缺点。此外，智能控制功能往往具有多种，无论用户是否需要，这些控制功能在设备出厂时通常都已经固定在设备中实现，且无法再增加其他的功能。不必要的功能加大了用户的投资，且由于所运行程序的庞大，会导致设备运行效率低下；而新的功能升级也可能会导致设备的软硬件升级、改动较大。
图1A和图1B示出了现有的两种智能路灯。其中，图1A所示的智能路灯可执行微基站1001、热点覆盖1002、环境监测1003、安防监控1004、广告运营1005、户外广播1006、充电管理1007和多功能网关1008多种控制应用。图1B所示的智能路灯可执行WIFI网络1011、环境监测1012、视频监控1013、信息发布1014、智能照明1015、网络基站1016、城市广播1017和汽车充电1018多种控制应用。上述两种智能路灯的控制应用设置固定，缺乏灵活性。
本发明实施例提供了一种控制装置，其具有智能控制功能，通过采用人工智能的算法实现对设备的智能控制，提高了设备控制的灵活性。其附加在传统设备上，无需替换传统设备，且其控制功能可根据用户需要灵活配置。其人工智能算法和控制逻辑的实现能够灵活变更，无需设备做出较大改动。
在本发明的一些实施例中，云平台可根据用户需求确定对设备的控制应用，并确定实现该控制应用所需的微服务。控制装置从云平台处获取该控制应用的逻辑和实现该控制应用所需的微服务。控制装置的AI芯片加载微服务，在运行微服务时采集传感器数据并据此进行AI推断；而控制装置的处理器按照该控制应用的逻辑，基于AI推断的结果生成设备的控制信号。由于控制应用可灵活配置，无需加载不需要的微服务，且可容易地加载新的微服务，实现了对设备的灵活控制。
在本发明的另一些实施例中，控制装置具有一个控制开关，操作人员可设置该控制开关以配置该控制装置所实现的控制应用。各种控制应用的逻辑和实现控制应用所需的微服务实现存储在控制装置的存储器中，根据该控制开关的配置从存储器中获取所需的逻辑和微服务，通过运行微服务得到AI推断的结果，并运行逻辑生成控制信号，来实现对设备的灵活控制。
在本发明的一些示例中，以路灯的控制装置为例，说明了微服务的加载，控制应用逻辑的运行，以及控制信号的生成与传递。并以具体的控制应用为例，说明了本发明实施例所实现的控制装置在实现智能控制功能时的灵活性。
为了使本发明实施例的目的、技术方案和优点更加清楚明白，以下参照附图对本发明实施例进一步详细说明。其中，后续描述的实施例仅仅是本发明实施例的一部分，而非全部的实施例。
图2为本发明实施例提供的智能控制系统的结构示意图。如果2所示，该控制系统100用于实现对设备20的智能控制，包括：
一个控制装置10，被配置为生成设备20的控制信号。控制装置10可安装在设备20附近，比如图6中的路灯控制装置50采用了抱杆的安装方式，安装在路灯的灯杆上。控制装置10通过自身的传感器104采集传感器数据，比如温度传感器用于采集设备20附近的温度，湿度传感器用于采集设备20附近的湿度，摄像头用于拍摄设备20附近环境的图片等。控制装置10基于采集的传感器数据进行AI推断，进而生成设备20的控制信号。
与以往智能控制装置不同的是，本发明实施例中，控制装置10所实现的控制功能可以灵活配置。控制装置10从该控制系统100中的云平台30处获取用户所需控制应用的逻辑以及实现逻辑所需的至少一项微服务，控制装置10加载并运行微服务以获取传感器数据并据此进行AI推断。控制装置10基于AI推断的结果按照获取的该控制应用的逻辑处理后，生成设备20的控制信号。由于控制应用的逻辑和微服务都是可以灵活配置的，因此实现了对设备20的灵活控制。
控制装置10生成的控制信号可经由云平台30发送至设备20的控制服务器40，并由该控制服务器40经过转换后发送至设备20的控制器，实现对设备20的控制。或者，控制装置10生成的控制信号可发送至控制服务器40，并由该控制服务器40经过变换后发送至设备服务器(30)的控制器，实现对设备20的控制。再或者，控制装置10生成的控制信号直接发送至设备20的控制器，实现对设备20的控制。由于不同设备的读取信息的格式不同，所以在控制信号的转发过程中，可能需要对控制信号的格式进行变换。此外，控制装置10生成的控制信号可能是用于控制一个区域内的所有设备20，比如在一个区域内部分设备20上部署控制装置10，生成的控制信号可用于对该区域内的所有设备20进行控制。此情况下，控制服务器40将控制信号变换成针对每一个设备20的信号并逐一发送。其中控制装置10可采用无线通信的方式与云平台30或者设备20的控制器通信，实现控制信号的传递。可采用的无线通信方式包括但不限于WiFi、第四代移动通信4G、LoRa等。
采用图2所示的控制系统100，无需用具有智能控制功能的新设备替换传统设备，且控制功能的设置灵活。并且，当控制应用的逻辑，或者微服务进行的AI推断的模型或判决门限、时序关系等有变化时，也可通过加载更新后的逻辑和微服务，容易地实现控制应用的更新。
图3为本发明实施例提供的智能控制方法的流程图。如图3所示，该方法可包括如下步骤：
S300：控制系统100中的云平台30确定需要实现的对设备20的一种控制应用。其中，控制系统100可根据用户需求确定控制应用。比如：用户需要实现路灯亮度的自动调节和车流量的监控，则控制系统100确定控制应用包括：亮度调节应用和车流量监控应用。
每一种控制应用是按照一定的逻辑实现的，并且需要基于一些AI推断的结果，而本发明实施例中，AI推断采用了微服务的方式实现。比如：亮度调节应用需要的微服务包括能见度计算微服务和路况判断微服务，分别用于确定路灯附近路段的能见度和路灯附近路段路况的判断；而车流量监控应用需要的微服务包括交通流量微服务，用于确定路灯附近路段的实时交通流量。以下面简单的例子来说明亮度调节应用的逻辑，比如：按照路灯附近路段的能见度和路灯附近路段路况来确定路灯的亮度，当能见度计算微服务进行AI推断的结果是路灯附近路段的能见度小于第一能见度阈值，且路况判断微服务进行AI推断的结果确定路灯附近路段出现了微天气时，则确定路灯的亮度为第一亮度；当能见度计算微服务进行AI推断的结果是路灯附近路段的能见度不小于第一能见度阈值，且路况判断微服务进行AI推断的结果确定路灯附近路段未出现微天气时，则确定路灯的亮度为第二亮度，其中第二亮度小于第一亮 度，……。
S301：控制装置10从云平台30处获取微服务及控制应用的逻辑。
其中，控制装置10将云平台30所确定的控制应用的逻辑以及实现该逻辑所需的各项微服务的代码下载到本地。控制装置10可采用无线通信的方式与云平台30建立无线连接并借此下载代码。
S302：控制装置10加载各项微服务。
S303：控制装置10运行各项微服务。在每一项微服务的运行过程中，获取该项微服务运行所需的传感器数据，并基于获取的传感器数据执行该项微服务的AI推断。基于传感器数据执行AI推断的示例可参见图7、图8及对应的描述。需要说明的是，微服务在运行过程中可依据从传感器处采集的数据进行AI推断，也可基于其他微服务进行AI推断的结果进行进一步的AI推断。
S304：控制装置10根据各项微服务的AI推断的结果，按照上述控制应用的逻辑生成用于控制设备20的控制信号。
S305：控制装置10将控制信号发送至设备20。
其中步骤S305有包括下列实现方式在内的多种方式：
方式一、控制装置10将控制信号直接发送至设备20；
方式二、控制装置10将控制信号发送至云平台30，而云平台30将控制信号发送至控制服务器40，控制服务器40再将控制信号发送至设备20的控制器，从而实现对设备20的控制。
此外，云平台30还可利用自身的运算资源对各项微服务进行更新。其中，可利用一项微服务执行AI推断时所需传感器数据的历史数据进行训练，得到更新后的数学模型、判决门限、时序关系等，控制装置10可定期或在微服务有更新时从云平台30处获取更新后的微服务。微服务的更新能够进一步保证AI推断结果的准确性。
图4为本发明实施例提供的智能控制装置的结构示意图。如图4所示，控制装置10可包括：
至少一个传感器104，被配置为采集传感器数据。其中传感器104在硬件结构上可位于控制装置10的内部，可以位于控制装置10的外部，与控制装置10之间通过无线通信方式传递传感器数据。传感器104采集设备20的数据，比如温度、湿度、电机转速、液体流速度等，采集到的传感器数据用于后续的AI推断。
通信模块106，被配置为与控制装置10之外的其他设备进行通信，比如：从云平台30处获取微服务以及获取的微服务所实现的控制应用的逻辑。如前所述，一项微服务用于实现一种AI推断，而云平台30根据用户的需求确定控制装置10要执行的控制应用，以及控制应用实现时所需的各项微服务，而控制装置10的通信模块106负责从云平台30处获取要执行的控制应用的逻辑及各项微服务的代码。其中，控制装置10可执行一种控制应用，也可执行多种控制应用，执行哪些控制应用由云平台30决定，控制装置10仅需获取逻辑和微服务的代码即可。
人工智能AI芯片101，被配置为加载通信模块104所获取的各项微服务，运行各项微服务。在一项微服务的运行过程中，从传感器104处获取该项微服务运行所需的传感器数据并基于获取的传感器数据执行该项微服务的AI推断。
处理器102，被配置为基于一个控制应用执行所需的各项微服务中的每一项执行AI推断的结果，并按照该控制应用的逻辑生成用于控制设备20的控制信号。
这里AI推断和控制应用逻辑分别由AI芯片和处理器执行的好处在于，AI芯片通常具有较强的运算能力，能够满足AI推断对运算能力的要求，实时性好，而处理器可用于逻辑的控制，与各个外围设备，比如通信模块等接口。
此外，通信模块104，还被配置为在处理器102的控制下从云平台30处获取更新后的微服务。其中，云平台30根据一项微服务在执行AI推断时所需的传感器数据的历史数据进行训练从而更新AI推断时使用的数据模型、判决门限、时序关系等。而AI芯片101，还被配置为加载更新后的第一微服务，并运行更新后的第一微服务进行AI推断。其中，通信模块104可定期从云平台30处获取更新后的微服务，也可在收到云平台30的更新通知后获取更新后的微服务。
图5为本发明实施例提供的智能控制装置的又一结构示意图。与图4所示结构不同的是，图5所示的控制装置10还包括一个控制开关105，被配置为设置需要对设备20执行的控制应用，比如：亮度调节应用、车流量监控应用等。而所有控制应用的逻辑以及每一种控制应用实现时所需的微服务的代码均预先存储在控制装置10的存储器103中，或者通过通信模块10预先从云平台30处获取并存储在存储器103中。操作人员通过设置该控制开关105，设置需要控制装置10执行的控制应用，处理器102读取该控制开关105设置的值，确定需要控制装置10执行的控制应用，并从存储器103中存储的多个控制应用逻辑的代码中找到确定的控制应用的代码，而AI芯片101从存储器103中存储的多项微服务的代码中找到实现该控制应 用所需的微服务的代码，加载并运行这些微服务。
下面以路灯的控制装置作为示例加以说明。图6为本发明实施例提供的智能路灯控制装置安装在传统路灯上的示意图。如图6所示，路灯控制装置50采用抱杆的方式安装在传统路灯的灯杆上，包括湿度传感器5041、温度传感器5042和摄像头5043以及雷达5044。其中，各个传感器可采用即插即用的方式，根据用户需求灵活配置。路灯控制装置50的处理器502从云平台80处获取控制应用的逻辑的代码以及实现该控制应用所需的微服务的代码5，路灯控制装置50的AI芯片501记载微服务的代码，运行微服务以读取各传感器的数据并进行AI推断。AI推断的结果1被发送至处理器502，由处理器502按照之前获取的逻辑进行处理后，得到路灯控制信号2或路灯控制信号7。其中，路灯控制信号2被发送至云平台80，再经过云平台80处理后形成路灯控制信号3，送至路灯的路灯控制服务器70处，比如路灯所在路段中所有路灯所共同使用的路灯控制服务器70。路灯控制服务器70将路灯控制信号3转换成控制用于控制路灯的路灯控制信号4，发送至路灯的控制器60。另一种实现方式是，路灯控制信号7由路灯控制装置50直接发送至路灯的控制器60，用于实现对路灯的控制。在硬件上，传感器可根据用户需求实现即插即用，在软件上，各项微服务可灵活地加载运行和停用。
图7为本发明实施例提供的智能路灯控制装置中AI芯片上加载的各项微服务的示意图。如图7所示，AI芯片101上加载有如下微服务：
能见度计算微服务5011，用于根据摄像头5043采集的图片或视频计算路灯附近空气能见度；
路况判断微服务5012，用于根据摄像头5043采集的图片或视频进行AI推断，确定路灯附近的路况，比如：冰雪路面、路面上有障碍物等；
行人密度微服务5013，用于根据摄像头5043采集的图片或视频计算路灯附近的行人的密度；
交通流量微服务5014，用于根据摄像头5043采集的图片或视频计算路灯附近的交通流量，或者根据雷达5044扫描的结果计算路灯附近的交通流量，再或者根据摄像头5043采集的图片或视频以及雷达5044扫描的结果计算路灯附近的交通流量；
微天气判断微服务5015，用于根据湿度传感器5041采集的路灯附近空气湿度、温度传感器5042采集的路灯附近温度以及能见度计算微服务5011计算得到的路灯附近空气能见度、 路况判断微服务5012推断得到的路等附近的路况信息来判断是否发生了微天气的情况，比如：下雨、下雪、有雾等；
特殊事件判断微服务5016，用于根据交通流量微服务5014计算得到的路灯附近的交通流量、行人密度微服务5013计算得到的路灯附近的行人密度，以及云平台80基于训练得到的行人密度和交通流量的正常模式6，判断路灯附近是否发生了特殊事件，比如：长时间的交通拥堵(交通拥堵的持续时长超过预设的时长阈值)等。其中，云平台80可基于雷达5044和/或摄像头5043所采集数据的历史值来训练得到正常模式6。或者云平台80也可基于交通流量微服务5014计算得到的路灯附近交通流量的历史值以及行人密度微服务5013计算得到的路灯附近的行人密度的历史值来训练得到正常模式6。
微天气判断微服务5015以及特殊事件判断微服务5016进行AI推断的结果1被发送至处理器502以用于生成路灯控制信号2。
图8示出了本发明实施例提供的路灯控制装置可加载的几种微服务的示例。其中不同微服务可适用于不同场景，能够根据相应的逻辑控制路灯的灯光亮度，这些微服务可根据用户需求灵活修改，而对应的传感器也可采用即插即用的方式，实现硬件的灵活配置。
1)昼夜时间设置微服务5017
在不同的季节，不同的城市，夜间的起止时间不同。这些数据可由云平台80或路灯控制装置50从气象机构或网站预先获取，并预先存储在路灯控制装置50中。如果有变更请求(例如政府希望路灯在黑暗时间之前亮起)，云平台80也可以年或月为周期更新该项微服务。
2)交通流量微服务5014和/或行人密度检测5013
在午夜，当道路上，特别是在高速公路上，车辆或行人很少时，可通过将路灯的亮度调到较低水平，节省大量的能源和成本。调整可以基于每周的策略，比如：分析交通流量的行人密度的历史数据以确定一个路灯附近交通流量和行人密度的正常模式。调整也可基于更灵活的模式，比如：例如，在路灯的灯光下观察到车辆或行人时，将路灯的亮度调高，以尽量减少潜在的事故、降低犯罪率。可以周为周期更新微这两项微服务。
3)微天气判断微服务5015
与传统天气相比，微天气对汽车事故的影响往往更为深远。比如在高速公路上，当湿度高且光线强时，可能会产生一种称为辐射雾的微天气，这种微天气非常厚，有时会沿着高速公路移动，导致驾驶员几乎没有时间做出反应。因此，如果可以加载微天气判断微服务5015，可使路灯附近道路更亮，也可作为对汽车驾驶员的警告，在很大程度上可提高道路安全水平。 可以天为周期更新该项微服务。
4)特殊事件判断微服务5016
路灯控制装置50还可加载特殊事件判断微服务5016。例如，在每年的长假期间，假期临近结束时通常会出现交通堵塞，在这种特殊情况下，路灯的照明可能需要保持更长时间，且亮度更高，但在假期期间，某些路段可能比平时的交通流量少很多，此时可控制路灯切换到更节能的模式。可以小时更新该项微服务。
图8中间示出了可灵活加载的上述几项微服务，其左侧相应地示出了运行一项微服务所需配置的传感器类型，其中最上方表示气象机构或网站的服务器，昼夜时间设置微服务5017在运行过程中可从服务器获取路灯所在城市不同日期期间的夜间起止时间。图8所示微服务右侧的柱状图示出了基于对应的微服务AI推断的结果计算得到的一天中不同时间段的路灯亮度。
此外，本发明实施例还提供一种计算机可读介质，该计算机可读介质存储有计算机可读代码，计算机可读代码用于对一个待控制的设备(比如：路灯)进行控制。该计算机可读代码可包括：
至少一项微服务的代码，其中，一项微服务用于实现一种人工智能AI推断，当至少一项微服务被加载到一个AI芯片上时，AI芯片运行至少一项微服务中的每一项，以获取该项微服务运行所需传感器数据并基于获取的传感器数据执行AI推断；
至少一项微服务实现的一种控制应用的逻辑的代码，当逻辑的代码被一个处理器执行时，基于至少一项微服务中的每一项执行AI推断的结果，并按照逻辑生成控制带控制设备的控制信号。
计算机可读介质的实施例包括软盘、硬盘、磁光盘、光盘(如CD-ROM、CD-R、CD-RW、DVD-ROM、DVD-RAM、DVD-RW、DVD+RW)、磁带、非易失性存储卡和ROM。可选地，可以由通信网络从服务器计算机上或云上下载计算机可读指令。计算机可读介质的一个例子是云平台30上的存储设备。该计算机可读介质的另一个例子是控制装置10的存储器103。
需要说明的是，上述各流程和各系统结构图中不是所有的步骤和模块都是必须的，可以根据实际的需要忽略某些步骤或模块。各步骤的执行顺序不是固定的，可以根据需要进行调整。上述各实施例中描述的系统结构可以是物理结构，也可以是逻辑结构，即，有些模块可能由同一物理实体实现，或者，有些模块可能分由至少两个物理实体实现，或者，可以由至 少两个独立设备中的某些部件共同实现。
以上各实施例中，硬件单元可以通过机械方式或电气方式实现。例如，一个硬件单元可以包括永久性专用的电路或逻辑(如专门的处理器，现场可编程门阵列(Field-Programmable Gate Array，FPGA)或专用集成电路(Application Specific Integrated Circuits，ASIC)等)来完成相应操作。硬件单元还可以包括可编程逻辑或电路(如通用处理器或其它可编程处理器)，可以由软件进行临时的设置以完成相应操作。具体的实现方式(机械方式、或专用的永久性电路、或者临时设置的电路)可以基于成本和时间上的考虑来确定。
上文通过附图和优选实施例对本发明实施例进行了详细展示和说明，然而本发明实施例不限于这些已揭示的实施例，基于上述实施例本领域技术人员可以知晓，可以组合上述不同实施例中的代码审核手段得到本发明更多的实施例，这些实施例也在本发明实施例的保护范围之内。

Claims (10)

  1. 控制系统(100)，用于对一个设备(20)进行控制，其特征在于，包括：
    一个云平台(30)，被配置为：
    确定需要实现的对所述设备(20)的一种控制应用；
    确定实现所述控制应用所需的至少一项微服务，其中，一项微服务用于实现一种人工智能AI推断；
    确定所述至少一项微服务实现所述控制应用的逻辑；
    一个控制装置(10)，被配置为：
    从所述云平台(30)处获取所述至少一项微服务以及所述逻辑；
    加载所述至少一项微服务；
    运行所述至少一项微服务中的每一项，以获取该项微服务运行所需传感器数据并基于获取的传感器数据执行该项微服务的AI推断；
    基于所述至少一项微服务中的每一项执行AI推断的结果，并按照所述逻辑生成用于控制所述设备(20)的控制信号。
  2. 如权利要求1所述的控制系统(100)，其特征在于，
    所述云平台(30)，还被配置为：针对所述至少一项微服务中的第一微服务，基于所述第一微服务执行AI推断时所需的传感器数据的历史数据进行训练，以更新所述第一微服务；
    所述控制装置(10)，还被配置为：从所述云平台(30)处获取更新后的所述第一微服务，并加载更新后的所述第一微服务。
  3. 如权利要求1或2所述的控制系统(100)，其特征在于，
    所述控制装置(10)还被配置为将所述控制信号发送至所述设备(20)；或者
    所述控制装置(10)还被配置为将所述控制信号发送至所述云平台(30)，且所述云平台(30)还被配置为将所述控制信号发送至用于控制所述设备(20)的控制服务器(40)。
  4. 控制装置(10)，用于对一个设备(20)进行控制，其特征在于，包括：
    至少一个传感器(104)，被配置为采集传感器数据；
    一个通信模块(106)，被配置为从一个云平台(30)处获取至少一项微服务以及所述至少一项微服务所实现的一种控制应用的逻辑，其中，一项微服务用于实现一种人工智能AI推断；
    一个人工智能AI芯片(101)，被配置为：加载所述至少一项微服务，运行所述至少一项微服务中的每一项，以获取该项微服务运行所需传感器数据并基于获取的传感器数据执行该项微服务的AI推断；
    一个处理器(102)，被配置为基于所述至少一项微服务中的每一项执行AI推断的结果，并按照所述逻辑生成用于控制所述设备(20)的控制信号。
  5. 如权利要求4所述的控制装置(10)，其特征在于，
    所述通信模块(106)，还被配置为获取更新后的第一微服务，其中所述第一微服务为所述至少一项微服务中的一项，且更新后的所述第一微服务是基于所述第一微服务执行AI推断时所需的传感器数据的历史数据进行训练得到的；
    所述AI芯片(101)，还被配置为加载所述更新后的所述第一微服务。
  6. 控制方法，用于对一个设备(20)进行控制，其特征在于，包括：
    获取至少一项微服务以及所述至少一项微服务实现的一种控制应用的逻辑，其中一项微服务用于实现一种人工智能AI推断；
    加载所述至少一项微服务；
    运行所述至少一项微服务中的每一项，以获取该项微服务运行所需传感器数据并基于获取的传感器数据执行AI推断；
    基于所述至少一项微服务中的每一项执行AI推断的结果，并按照所述逻辑生成用于控制所述设备(20)的控制信号。
  7. 如权利要求6所述的方法，其特征在于，还包括：
    获取更新后的第一微服务，其中所述第一微服务为所述至少一项微服务中的一项，且更新后的所述第一微服务是基于所述第一微服务执行AI推断时所需的传感器数据的历史数据进行训练得到的；
    加载更新后的所述第一微服务。
  8. 计算机可读介质，其特征在于，所述计算机可读介质存储有计算机可读代码，所述计算机可读代码用于对一个设备(20)进行控制，包括：
    至少一项微服务的代码，其中，一项微服务用于实现一种人工智能AI推断，当所述至少 一项微服务被加载到一个AI芯片(101)上时，所述AI芯片(101)运行所述至少一项微服务中的每一项，以获取该项微服务运行所需传感器数据并基于获取的传感器数据执行AI推断；
    所述至少一项微服务实现的一种控制应用的逻辑的代码，当所述逻辑的代码被一个处理器(102)执行时，基于所述至少一项微服务中的每一项执行AI推断的结果，并按照所述逻辑生成控制所述设备(20)的控制信号。
  9. 控制装置(10)，用于对一个设备(20)进行控制，其特征在于，包括：
    至少一个传感器(104)，被配置为采集传感器数据；
    一个存储器(103)，被配置为存储复数项微服务的代码，以及至少一种控制应用中每一种控制应用逻辑的代码，其中，一项微服务用于实现一种人工智能AI推断，实现每一种控制应需要所述复数项微服务中的部分或全部；
    一个控制开关(105)，被配置为设置需要对所述设备(20)进行的一种控制应用；
    一个处理器(102)，被配置为运行所述控制开关(105)所设置的控制应用逻辑的代码；
    一个人工智能AI芯片(101)，被配置为：
    从所述存储器(103)中加载实现所述控制开关(105)所设置的控制应用所需的至少一项微服务；
    运行所述至少一项微服务中的每一项，以获取该项微服务运行所需传感器数据并基于获取的传感器数据执行该项微服务的AI推断；
    所述处理器(102)，还被配置为基于所述至少一项微服务中的每一项执行AI推断的结果，并按照所述逻辑生成用于控制所述设备(20)的控制信号。
  10. 如权利要求9所述的控制装置(10)，其特征在于，还包括一个通信模块(106)，被配置为：
    将所述控制信号发送至所述设备(20)；或者
    将所述控制信号发送至一个云平台(30)，且所述云平台(30)还被配置为将所述控制信号发送至用于控制所述设备(20)的控制服务器(40)。
PCT/CN2019/096041 2019-07-15 2019-07-15 控制方法、装置、系统和计算机可读介质 WO2021007755A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN201980096053.9A CN113785270A (zh) 2019-07-15 2019-07-15 控制方法、装置、系统和计算机可读介质
PCT/CN2019/096041 WO2021007755A1 (zh) 2019-07-15 2019-07-15 控制方法、装置、系统和计算机可读介质

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2019/096041 WO2021007755A1 (zh) 2019-07-15 2019-07-15 控制方法、装置、系统和计算机可读介质

Publications (1)

Publication Number Publication Date
WO2021007755A1 true WO2021007755A1 (zh) 2021-01-21

Family

ID=74210114

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/096041 WO2021007755A1 (zh) 2019-07-15 2019-07-15 控制方法、装置、系统和计算机可读介质

Country Status (2)

Country Link
CN (1) CN113785270A (zh)
WO (1) WO2021007755A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114167759A (zh) * 2021-11-25 2022-03-11 中万恩科技有限公司 一种4g信号传输与数据管理的洗碗机控制器

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20150281002A1 (en) * 2013-09-09 2015-10-01 Tsinghua University Universal support system for internet of things
CN105404169A (zh) * 2015-12-29 2016-03-16 上海大学 智能家居中央控制系统及其控制方法
CN106533929A (zh) * 2016-12-30 2017-03-22 北京中电普华信息技术有限公司 一种微服务开发平台、生成方法以及部署方法和装置
CN108508856A (zh) * 2018-03-28 2018-09-07 西安西电电气研究院有限责任公司 一种工业设备智能控制系统及方法

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108573607A (zh) * 2017-03-10 2018-09-25 北京嘀嘀无限科技发展有限公司 一种红绿灯控制系统及方法
US10255052B2 (en) * 2017-05-30 2019-04-09 International Business Machines Corporation Dynamic deployment of an application based on micro-services

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20150281002A1 (en) * 2013-09-09 2015-10-01 Tsinghua University Universal support system for internet of things
CN105404169A (zh) * 2015-12-29 2016-03-16 上海大学 智能家居中央控制系统及其控制方法
CN106533929A (zh) * 2016-12-30 2017-03-22 北京中电普华信息技术有限公司 一种微服务开发平台、生成方法以及部署方法和装置
CN108508856A (zh) * 2018-03-28 2018-09-07 西安西电电气研究院有限责任公司 一种工业设备智能控制系统及方法

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114167759A (zh) * 2021-11-25 2022-03-11 中万恩科技有限公司 一种4g信号传输与数据管理的洗碗机控制器
CN114167759B (zh) * 2021-11-25 2023-06-16 中万恩科技有限公司 一种4g信号传输与数据管理的洗碗机控制器

Also Published As

Publication number Publication date
CN113785270A (zh) 2021-12-10

Similar Documents

Publication Publication Date Title
US10653014B2 (en) Systems and methods for an intermediate device structure
CN113409579B (zh) 一种基于ai物联网技术的智慧城市交通管控系统
US11250699B2 (en) System and method of adaptive traffic management at an intersection
JP5773546B2 (ja) 屋外照明システムのコントロール方法、装置、および設備
JP6629205B2 (ja) 隣接する照明器具及び／又は接続されたデバイスからのステータス情報に基づく適合検出設定を備えるセンサネットワーク
US20190096242A1 (en) Traffic light control device, method, and system
WO2018232906A1 (zh) 一种智能路灯监控方法及装置
CN112040154A (zh) 用于减少成像光源中的闪烁伪影的系统和方法
CN109383362B (zh) 一种汽车前照灯控制方法和装置
US10568188B2 (en) On-demand street lighting for a connected vehicle
US10523759B2 (en) Determining infrastructure lamp status using a vehicle
US6813554B1 (en) Method and apparatus for adding commercial value to traffic control systems
WO2016000325A1 (zh) 路灯控制方法及装置
CN101572008A (zh) 具有违规抓拍和交通疏导功能的红绿灯控制系统
US20190188503A1 (en) Vehicle monitoring of infrastructure lighting
WO2018227724A1 (zh) 一种路灯控制方法及装置
CN110874920A (zh) 一种路况监测方法及装置、服务器、电子设备
WO2021007755A1 (zh) 控制方法、装置、系统和计算机可读介质
KR20000058895A (ko) 유무선 통합 교통신호 제어시스템
US8081801B2 (en) Device and method for controlling streetlights
JP2003022493A (ja) 信号機システム、信号機を用いた情報提供収集方法、信号機を用いた情報提供収集プログラム及びコンピュータ読み取り可能な記録媒体
CN113115203B (zh) 一种路侧单元设备的降耗方法、装置及系统
US10979837B2 (en) Microphone calibration system
WO2019142697A1 (ja) 情報収集システム、制御装置およびプログラム
CN105448114A (zh) 一种智能交通路口信息系统

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: 19938087

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: 19938087

Country of ref document: EP

Kind code of ref document: A1