WO2021036456A1 - 基于大数据和人工智能的混合动力推进方法和机器人系统 - Google Patents

基于大数据和人工智能的混合动力推进方法和机器人系统 Download PDF

Info

Publication number
WO2021036456A1
WO2021036456A1 PCT/CN2020/097978 CN2020097978W WO2021036456A1 WO 2021036456 A1 WO2021036456 A1 WO 2021036456A1 CN 2020097978 W CN2020097978 W CN 2020097978W WO 2021036456 A1 WO2021036456 A1 WO 2021036456A1
Authority
WO
WIPO (PCT)
Prior art keywords
data
preset
model
power propulsion
type
Prior art date
Application number
PCT/CN2020/097978
Other languages
English (en)
French (fr)
Inventor
朱定局
Original Assignee
南京智慧光信息科技研究院有限公司
大国创新智能科技(东莞)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 南京智慧光信息科技研究院有限公司, 大国创新智能科技(东莞)有限公司 filed Critical 南京智慧光信息科技研究院有限公司
Publication of WO2021036456A1 publication Critical patent/WO2021036456A1/zh

Links

Images

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W20/00—Control systems specially adapted for hybrid vehicles
    • B60W20/10—Controlling the power contribution of each of the prime movers to meet required power demand
    • B60W20/11—Controlling the power contribution of each of the prime movers to meet required power demand using model predictive control [MPC] strategies, i.e. control methods based on models predicting performance
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning

Definitions

  • the invention relates to the field of information technology, in particular to a hybrid propulsion method and a robot system based on big data and artificial intelligence.
  • the hybrid propulsion methods in the prior art include electric propulsion, diesel-gasoline and other thermal propulsion methods, etc., or multiple energy hybrid propulsion methods;
  • the switching of the hybrid propulsion mode in the prior art generally adopts manual switching. It can be seen that the switching mode of the hybrid propulsion mode in the prior art is single and low in intelligence.
  • an embodiment of the present invention provides a hybrid propulsion method, the method including:
  • the object type obtaining step is used to obtain the type of the object to which the power propulsion device belongs;
  • the data acquisition step is used to acquire preset types of data related to the object
  • the preset model obtaining step is used to obtain the preset model
  • the model calculation step is used to input the preset type of data related to the object into the preset model, and the output calculated by the preset model is used as the recommended power propulsion mode.
  • the method further includes:
  • the system control step is used to control the power propulsion device according to the recommended power propulsion mode.
  • the preset type is preset by the user or obtained from the knowledge base;
  • the preset type of data includes data related to the choice of power propulsion mode
  • the data related to the object includes data of the object and environmental data of the object;
  • the preset type of data related to the object is the preset type of current data or recent data related to the object or data within a recent preset time period.
  • the step of obtaining the preset model includes:
  • the historical big data obtaining step is used to obtain valid historical big data of all objects of the type to which the object belongs; the historical big data includes big data collected so far;
  • the corresponding data acquisition step is used to acquire the preset types of data in the effective historical big data and the corresponding power propulsion mode
  • the deep learning model initialization step is used to initialize the deep learning model
  • the unsupervised training step is used to use the preset type of data in the effective historical big data as the input of the deep learning model, and perform unsupervised training on the deep learning model;
  • the supervised training step is used to use the preset types of data in the historical big data and their corresponding power propulsion modes as the input and output of the deep learning model, respectively, and to determine the depth after unsupervised training.
  • Supervised training of learning model
  • the preset model generation step is used to obtain the deep learning model after supervised training as the preset model.
  • the step of obtaining the preset model includes:
  • the historical big data obtaining step is used to obtain valid historical big data of all objects of the type to which the object belongs;
  • the corresponding data acquisition step is used to acquire the preset types of data in the effective historical big data and the corresponding power propulsion mode
  • the model data setting step is used to use the preset type of data and the corresponding power propulsion mode in the effective historical big data as the to-be-matched data of the preset model and the corresponding to-be-recommended data respectively;
  • the model calculation steps include:
  • the matching step is used to perform fuzzy matching between the data related to the object of the preset type and each of the to-be-matched data in the preset model;
  • the recommendation step is used to obtain the recommended data corresponding to the selected data to be matched from the preset model as the output calculated by the preset model, and use the output as the recommended power propulsion mode.
  • the step of acquiring historical big data includes:
  • the manned control data acquisition step is used to obtain historical data of all similar objects that are "manned” and “have at least one common power propulsion method with the object" as the first big data;
  • Unmanned control data acquisition step for acquiring the history of all objects of the same type that are "unmanned” and “the selection effect of the power propulsion mode meets the preset conditions” and “have at least one power propulsion mode in common with the object” Data as the second largest data;
  • the unmanned data cleaning step is used to delete the power propulsion mode and other corresponding data that the object does not have from the second largest data to obtain the fourth largest data;
  • the effective historical big data generation step is used to use the third and fourth largest data as effective historical big data.
  • the object is a vehicle; the vehicle includes an unmanned vehicle; the preset type of data related to the object includes road condition data of the road section where the vehicle is currently located, and exhaust gas of the road section where the vehicle is currently located. Pollution control index data, noise control index data of the road section where the vehicle is currently located, the speed limit range of the road section where the vehicle is currently located, the remaining energy data of the vehicle's current different power propulsion types, the vehicle model, the One or more of the current preset data of the car.
  • the object is a ship; the ship includes an unmanned ship; the preset type of data related to the object includes the sea condition data of the current segment of the ship, and the current segment of the ship.
  • the object is an airplane; the airplane includes a drone; the preset type of data related to the object includes weather data of the flight segment where the aircraft is currently located, and data about the flight segment where the aircraft is currently located.
  • the noise pollution control index data the wind data of the current flight segment, the remaining energy data of the current different power propulsion types of the aircraft, the model, and the current preset data of the aircraft.
  • an embodiment of the present invention provides a system, wherein the system executes the steps in the hybrid propulsion method according to any one of the first aspect; the system includes a robot system.
  • the embodiment of the present invention obtains a preset model by learning from historical big data, and then calculates the power propulsion method that should be used currently through the preset model and current data, and the historical data and current data include the information of the object to which the power propulsion system belongs. Data and environmental data, so that the obtained preset models and recommended power propulsion methods are more in line with the needs of the object and the environment, and are more efficient. Therefore, the embodiment of the present invention can make the switching of power propulsion methods more intelligent and efficient. For example, for the switching of hybrid propulsion of an unmanned ship, the data that needs to be considered include the sea state data of the ship's current flight segment, the exhaust pollution control index data of the current flight segment, and the noise of the current flight segment.
  • control index data weather data of the current flight segment, wind data of the current flight segment, current other demand data, current remaining energy data of different power propulsion types, ship type, current other preset data, etc. Or several. Wherein, the flight segment can be replaced with sea area. Other current preset data include control index data for noise, speed, etc. during combat.
  • the prior art can only switch manually or automatically switch based on speed. Manual switching requires users to be highly demanding and unintelligent. If automatic switching based on speed only takes into account the speed factor, it is not suitable for a wider range of application scenarios. In the application scenario, not only the speed should be considered, but also more environmental factors and factors of the object of the power propulsion system should be considered.
  • the hybrid propulsion method and robot system based on big data and artificial intelligence provided by the embodiments of the present invention include: obtaining the type of the object to which the power propulsion device belongs, obtaining the data related to the object of the preset type, and obtaining the preset model , Input the preset type of data related to the object into the preset model, and the output calculated by the preset model is used as the recommended power propulsion mode.
  • the above method and system improve the intelligence and efficiency of hybrid propulsion mode switching through hybrid propulsion technology based on big data and artificial intelligence.
  • FIG. 1 is a flowchart of a hybrid propulsion method provided by Embodiment 2 of the present invention.
  • Embodiment 5 of the present invention is a flowchart of the steps of obtaining a preset model provided by Embodiment 5 of the present invention.
  • Fig. 5 is a schematic block diagram of a hybrid propulsion system provided by Embodiment 11 of the present invention.
  • FIG. 6 is a functional block diagram of a preset model module provided by Embodiment 13 of the present invention.
  • FIG. 7 is a functional block diagram of a preset model module provided by Embodiment 14 of the present invention.
  • FIG. 8 is a functional block diagram of a model calculation module provided by Embodiment 14 of the present invention.
  • a hybrid propulsion method includes an object type acquisition step S100, a data acquisition step S200, a preset model acquisition step S300, a model calculation step S400, and a system control step S500.
  • the object type obtaining step S100 is used to obtain the type of the object to which the power propulsion device belongs.
  • the power propulsion device is the power propulsion device of the object.
  • the power propulsion device is installed on the object, so the object is the object to which the power propulsion device belongs.
  • the objects include vehicles, ships, airplanes, and other vehicles or other systems or equipment that need to be installed with power propulsion devices. Therefore, the power propulsion mode of the power propulsion device of the object of the type is determined according to the object type, and the preset type of data related to the object related to the propulsion mode is determined.
  • the data acquisition step S200 is used to acquire preset types of data related to the object.
  • the recommended power propulsion mode is calculated according to the preset type of data related to the object.
  • the preset model obtaining step S300 is used to obtain a preset model. Thereby, a corresponding relationship between the data related to the object of the preset type and the power propulsion mode is established through the preset model.
  • the input format of the preset model is a preset type of data format related to the object
  • the output format is the data format of the power propulsion mode
  • the data format of the power propulsion mode can use a digital format, and each power is propelled
  • the method is coded as a number
  • the power propulsion method that is, the power propulsion method
  • the power propulsion method includes the use of electric energy for propulsion, diesel engine for propulsion, gas turbine for propulsion, etc., and one or more propulsion methods mixed;
  • the model calculation step S400 is used to input the preset type of data related to the object into the preset model, and the output calculated by the preset model is used as the recommended power propulsion mode. Therefore, a recommended power propulsion method is provided for controlling the power propulsion system.
  • the output is the output of the preset model
  • the method according to embodiment 1 further includes a system control step S500, as shown in FIG. 1.
  • System control step S500 is used to control the power propulsion device according to the recommended power propulsion mode; thereby enabling the power propulsion system to operate in a better manner, thereby improving the intelligence of the power propulsion system of the object Sex and efficiency. Specifically, it is used to determine whether the recommended power propulsion mode is consistent with the current power propulsion mode: if not, a control instruction for switching the current power propulsion mode to the recommended power propulsion mode is sent to the power propulsion device.
  • the preset type is preset by the user or obtained from the knowledge base;
  • the preset type of data includes data related to the choice of power propulsion mode
  • the data related to the object includes data of the object and environmental data of the object;
  • the preset type of data related to the object is the preset type of current data or recent data related to the object or data within a recent preset time period.
  • the preset model acquisition step S300 includes historical big data acquisition step S311, corresponding data acquisition step S312, deep learning model initialization step S313, unsupervised training step S314, supervised training step S315, and preset model generation step S316, as shown in the figure 2 shown.
  • Historical big data obtaining step S311 is used to obtain valid historical big data of all objects of the type to which the object belongs;
  • Corresponding data acquisition step S312 is used to acquire a preset type of data related to the object and its corresponding power propulsion mode in the effective historical big data, wherein the preset type of data related to the object Historical data of the preset type;
  • the deep learning model initialization step S313 is used to initialize the deep learning model
  • the unsupervised training step S314 is used to use the preset type of data related to the object in the effective historical big data as the input of the deep learning model, and perform unsupervised training on the deep learning model;
  • the supervised training step S315 is used to use the preset type of data related to the object in the historical big data and the corresponding power propulsion mode as the input and output of the deep learning model, respectively. Performing supervised training on the deep learning model after supervised training;
  • the preset model generation step S316 is used to obtain the deep learning model after supervised training as the preset model.
  • the historical big data can be obtained online through the network or obtained from a large historical database.
  • the effective historical big data includes preset types of data related to the object and the corresponding power propulsion mode; the effective historical big data is collected in the past for a long period of time.
  • the preset type of data related to the object and its corresponding power propulsion mode are collected by each collected object at each time of collection. It is necessary to collect each collected object at each time of collection.
  • the preset type of data related to the object at the time of collection must also collect the power propulsion mode of each collected object at each of the collected moments, wherein each of the collected objects
  • the object belongs to a collection of objects of the type to which the object belongs.
  • the type of the object includes objects (including vehicles) that have at least one power propulsion mode in common with the object.
  • the preset model obtaining step S300 includes historical big data obtaining step S321, corresponding data obtaining step S322, and model data setting step S323, as shown in FIG. 3.
  • Historical big data acquisition step S321 is used to acquire valid historical big data of all objects of the type to which the object belongs;
  • Corresponding data acquisition step S322 is used to acquire a preset type of data related to the object and its corresponding power propulsion mode in the effective historical big data; wherein, the preset type of data related to the object Historical data of the preset type;
  • the model data setting step S323 is used to use the preset type of data related to the object and its corresponding power propulsion mode in the effective historical big data as the to-be-matched data of the preset model and its corresponding respectively The data to be recommended.
  • the model calculation step S400 includes a matching step S421, a selection step S422, and a recommendation step S423, as shown in FIG. 4.
  • the matching step S421 is used for the fuzzy matching of the data related to the object of the preset type with each of the data to be matched in the preset model;
  • Selecting step S422 is for selecting the data to be matched in the preset model that has the greatest degree of matching with the input data of the preset type and related to the object;
  • the recommendation step S423 is for obtaining the data to be recommended corresponding to the selected data to be matched from the preset model as the output calculated by the preset model, and using the output as the recommended power propulsion mode .
  • Embodiment 4 uses big data and deep learning technology
  • embodiment 5 uses big data and its recommendation technology.
  • the historical big data acquisition step S321 includes:
  • Manned control data acquisition step S321-1 used to obtain historical data of all similar objects (for example, cars) that are "manned (for example, manned driving)" and "have at least one common power propulsion mode with the object" as the first A large amount of data;
  • the object of the same type is an object of the type to which the object belongs;
  • the object of the same type and the object belong to the type of the object;
  • the historical data includes the data of the preset type and its corresponding power Propulsion method; because humans are intelligent, the power propulsion method selected by the object when manned (for example, manned driving) is more credible, while unmanned control (for example, unmanned The historical big data of the object (for example, car) of driving) is not necessarily credible.
  • the manned control includes manned selection of power propulsion mode;
  • Unmanned control data acquisition step S321-2 Obtain “unmanned control (for example, unmanned driving)" and “the selection effect of the power propulsion mode satisfies the preset conditions” and “has at least one power propulsion mode in common with the object
  • the user score including the selection effect of the power propulsion mode is greater than the preset threshold.
  • the unmanned control includes unmanned selection of power propulsion mode;
  • Manned data cleaning step S321-3 Delete the power propulsion mode that the object does not have and other corresponding data from the first big data, including deleting the power propulsion mode that the object does not have from the first big data and The corresponding data of the preset type obtains the third largest data;
  • Unmanned data cleaning step S321-4 Delete the power propulsion method that the object does not have and other corresponding data from the second largest data, including deleting the power propulsion method that the object does not have from the second largest data And the corresponding data of the preset type to obtain the fourth largest data;
  • Effective historical big data generation step S321-5 Regard the third largest data and the fourth largest data as effective historical big data.
  • the object is a vehicle; the vehicle includes an unmanned vehicle;
  • the preset type of data related to the object includes road condition data of the road section where the vehicle is currently located, exhaust pollution control index data of the current road section, noise control index data of the current road section, and data on the current road section.
  • the track section can be replaced with a region.
  • Other current preset data include control index data for noise, speed, etc. during combat.
  • the object is a ship; the ship includes an unmanned ship;
  • the preset type of data related to the object includes the sea state data of the current flight segment of the ship, the exhaust pollution control index data of the current flight segment, the noise control index data of the current flight segment, and the current One or more of meteorological data of the flight segment, wind data of the current flight segment, current other demand data, current remaining energy data of different power propulsion types, ship type, current other preset data, and so on.
  • the flight segment can be replaced with sea area.
  • Other current preset data include control index data for noise, speed, etc. during combat.
  • the object is an airplane; the airplane includes a drone;
  • the preset type of data related to the object includes weather data of the flight segment where the aircraft is currently located, noise pollution control index data of the flight segment currently located, wind data of the currently located flight segment, and current different power propulsion One or more of the remaining energy data of the type, model, current other preset data, and so on.
  • the flight segment can be replaced with airspace.
  • Other current preset data include control index data for noise, speed, etc. during combat.
  • a hybrid propulsion system includes an object type acquisition module 100, a data acquisition module 200, a preset model acquisition module 300, and a model calculation module 400.
  • the object type acquiring module 100 is used to acquire the type of the object to which the power propulsion device belongs.
  • the data acquisition module 200 is configured to acquire preset types of data related to the object.
  • the preset model obtaining module 300 is used to obtain a preset model.
  • the model calculation module 400 is configured to input the preset type of data related to the object into the preset model, and the output calculated by the preset model is used as a recommended power propulsion mode.
  • the system according to the embodiment 10 further includes a system control module 500, as shown in FIG. 5.
  • the system control module 500 is used to control the power propulsion device according to the recommended power propulsion mode. Specifically, it is determined whether the recommended power propulsion mode is consistent with the current power propulsion mode: if not, a control instruction for switching the current power propulsion mode to the recommended power propulsion mode is sent to the power propulsion device.
  • the preset type is preset by the user or obtained from the knowledge base;
  • the preset type of data includes data related to the choice of power propulsion mode
  • the data related to the object includes data of the object and environmental data of the object;
  • the preset type of data related to the object is the preset type of current data or recent data related to the object or data within a recent preset time period.
  • the preset model acquisition module 300 includes a historical big data acquisition module 311, a corresponding data acquisition module 312, a deep learning model initialization module 313, an unsupervised training module 314, a supervised training module 315, and a preset model generation module 316, as shown in the figure 6 shown.
  • the historical big data acquisition module 311 is configured to acquire valid historical big data of all objects of the type to which the object belongs;
  • Corresponding data acquisition module 312 configured to acquire preset types of data related to the object and its corresponding power propulsion mode in the effective historical big data
  • the deep learning model initialization module 313 is used to initialize the deep learning model
  • the supervised training module 315 is configured to use the preset type of data related to the object in the historical big data and the corresponding power propulsion mode as the input and output of the deep learning model, respectively. Performing supervised training on the deep learning model after supervised training;
  • the preset model generation module 316 is configured to obtain the deep learning model after supervised training as the preset model.
  • the preset model acquisition module 300 includes a historical big data acquisition module 321, a corresponding data acquisition module 322, and a model data setting module 323, as shown in FIG. 7.
  • the historical big data acquisition module 321 is configured to acquire valid historical big data of all objects of the type to which the object belongs;
  • Corresponding data acquisition module 322 configured to acquire preset types of data related to the object and its corresponding power propulsion mode in the effective historical big data
  • the model data setting module 323 is configured to use the preset type of data related to the object and the corresponding power propulsion mode in the effective historical big data as the to-be-matched data of the preset model and the corresponding data respectively The data to be recommended.
  • the model calculation module 400 includes a matching module 421, a selection module 422, and a recommendation module 423, as shown in FIG. 8.
  • the matching module 421 is configured to perform fuzzy matching between the preset type of data related to the object and each of the to-be-matched data in the preset model;
  • the selecting module 422 is configured to select the data to be matched in the preset model that has the greatest degree of matching with the input data of the preset type and related to the object;
  • the recommendation module 423 is configured to obtain the recommended data corresponding to the selected data to be matched from the preset model as the output calculated by the preset model, and use the output as the recommended power propulsion mode .
  • the historical big data acquisition module 321 includes:
  • Manned control data acquisition module 321-1 used to acquire historical data of all similar objects (for example, cars) that are "manned (for example, manned driving)" and “have at least one common power propulsion mode with the object" as the first A large amount of data;
  • the object of the same type is an object of the type to which the object belongs;
  • the object of the same type and the object belong to the type of the object;
  • the historical data includes the data of the preset type and its corresponding power Way of advancing
  • Unmanned control data acquisition module 321-2 acquires "unmanned control (for example, unmanned driving)" and “the selection effect of the power propulsion mode satisfies the preset conditions” and “has at least one power propulsion mode in common with the object
  • Manned data cleaning module 321-3 Delete the power propulsion mode that the object does not have and other corresponding data from the first big data, including deleting the power propulsion mode that the object does not have from the first big data and The corresponding data of the preset type obtains the third largest data;
  • Unmanned data cleaning module 321-4 Delete the power propulsion method that the object does not have and other corresponding data from the second largest data, including deleting the power propulsion method that the object does not have from the second largest data And the corresponding data of the preset type to obtain the fourth largest data;
  • Effective historical big data generation module 321-5 Regard the third and fourth largest data as effective historical big data.
  • the system according to embodiment 10 includes:
  • the object is a vehicle; the vehicle includes an unmanned vehicle;
  • the preset type of data related to the object includes road condition data of the road section where the vehicle is currently located, exhaust pollution control index data of the current road section, noise control index data of the current road section, and data on the road section where the vehicle is currently located.
  • the track section can be replaced with a region.
  • Other current preset data include control index data for noise, speed, etc. during combat.
  • the object is a ship; the ship includes an unmanned ship;
  • the preset type of data related to the object includes the sea state data of the current flight segment of the ship, the exhaust pollution control index data of the current flight segment, the noise control index data of the current flight segment, and the current One or more of meteorological data of the flight segment, wind data of the current flight segment, current other demand data, current remaining energy data of different power propulsion types, ship type, current other preset data, and so on.
  • the flight segment can be replaced with sea area.
  • Other current preset data include control index data for noise, speed, etc. during combat.
  • the object is an airplane; the airplane includes a drone;
  • the preset type of data related to the object includes weather data of the flight segment where the aircraft is currently located, noise pollution control index data of the flight segment currently located, wind data of the currently located flight segment, and current different power propulsion One or more of the remaining energy data of the type, model, current other preset data, and so on.
  • the flight segment can be replaced with airspace.
  • Other current preset data include control index data for noise, speed, etc. during combat.
  • a robot system is provided, and the robots are respectively configured with the systems described in Embodiment 10 to Embodiment 18.
  • the methods and systems in the foregoing embodiments can be executed and deployed on computers, servers, cloud servers, supercomputers, robots, embedded devices, electronic devices, and so on.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Software Systems (AREA)
  • Medical Informatics (AREA)
  • Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • Mechanical Engineering (AREA)
  • Transportation (AREA)
  • Artificial Intelligence (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Automation & Control Theory (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Feedback Control In General (AREA)

Abstract

基于大数据和人工智能的混合动力推进方法和机器人系统，包括：获取动力推进装置所属的对象的类型，获取预设类型的与所述对象相关的数据，获取预设模型，将所述预设类型的与所述对象相关的数据输入所述预设模型，通过所述预设模型计算得到的输出作为推荐的动力推进方式。上述方法和系统通过基于大数据和人工智能的混合动力推进技术，提高了混合动力推进方式切换的智能性、高效性。

Description

基于大数据和人工智能的混合动力推进方法和机器人系统 技术领域
本发明涉及信息技术领域，特别是涉及一种基于大数据和人工智能的混合动力推进方法和机器人系统。
背景技术
在实现本发明过程中，发明人发现现有技术中至少存在如下问题：现有技术下混合动力推进方式包括电能推进方式、柴油汽油等热能推进方式、等等、或者多种能混合推进方式；现有技术下混合动力推进方式的切换一般采用手工切换，可见现有技术下混合动力推进方式的切换模式单一、智能性低。
因此，现有技术还有待于改进和发展。
发明内容
基于此，有必要针对现有技术中的缺陷或不足，提供基于大数据和人工智能的混合动力推进方法和机器人系统，以解决现有技术中混合动力推进方式切换的智能性低、高效性不足的缺点。
第一方面，本发明实施例提供一种混合动力推进方法，所述方法包括：
对象类型获取步骤，用于获取动力推进装置所属的对象的类型；
数据获取步骤，用于获取预设类型的与所述对象相关的数据；
预设模型获取步骤，用于获取预设模型；
模型计算步骤，用于将所述预设类型的与所述对象相关的数据输入所述预设模型，通过所述预设模型计算得到的输出作为推荐的动力推进方式。
优选地，所述方法还包括：
系统控制步骤，用于根据所述推荐的动力推进方式控制所述动力推进装置。
优选地，
所述预设类型由用户预先设置或从知识库中获取；
所述预设类型的数据包括与动力推进方式的选择有相关性的数据；
所述与所述对象相关的数据包括所述对象的数据、所述对象的环境数据；
所述预设类型的与所述对象相关的数据为所述预设类型的与所述对象相关 的当前数据或近期数据或最近预设时段内的数据。
优选地，所述预设模型获取步骤包括：
历史大数据获取步骤，用于获取所述对象所属类型的所有对象的有效历史大数据；所述历史大数据包括迄今为止所采集的大数据；
对应数据获取步骤，用于获取所述有效历史大数据中预设类型的数据及其对应的动力推进方式；
深度学习模型初始化步骤，用于初始化深度学习模型；
无监督训练步骤，用于将所述有效历史大数据中所述预设类型的数据作为所述深度学习模型的输入，对所述深度学习模型进行无监督训练；
有监督训练步骤，用于将所述历史大数据中所述预设类型的数据及其对应的动力推进方式分别作为所述深度学习模型的输入和输出，对通过无监督训练之后的所述深度学习模型进行有监督训练；
预设模型生成步骤，用于获取有监督训练之后的所述深度学习模型作为所述预设模型。
优选地，
所述预设模型获取步骤包括：
历史大数据获取步骤，用于获取所述对象所属类型的所有对象的有效历史大数据；
对应数据获取步骤，用于获取所述有效历史大数据中预设类型的数据及其对应的动力推进方式；
模型数据设置步骤，用于将所述有效历史大数据中所述预设类型的数据及其对应的动力推进方式分别作为所述预设模型的待匹配数据及其对应的待推荐数据；
所述模型计算步骤包括：
匹配步骤，用于所述将所述预设类型的与所述对象相关的数据与所述预设模型中的每一个所述待匹配数据进行模糊匹配；
选取步骤，用于选取与所述输入的所述预设类型的与所述对象相关的数据 匹配度最大的所述预设模型中的所述待匹配数据；
推荐步骤，用于从所述预设模型中获取与所选取的所述待匹配数据所对应的待推荐数据作为所述预设模型计算得到的输出，将所述输出作为推荐的动力推进方式。
优选地，所述历史大数据获取步骤包括：
有人控制数据获取步骤，用于获取“有人控制”且“与所述对象具有至少一种共同的动力推进方式”的所有同类对象的历史数据作为第一大数据；
无人控制数据获取步骤，用于获取“无人控制”且“动力推进方式的选择效果满足预设条件”且“与所述对象具有至少一种共同的动力推进方式”的所有同类对象的历史数据作为第二大数据；
有人控制数据清洗步骤，用于从第一大数据中删除所述对象不具备的动力推进方式及其对应的其他数据，得到第三大数据；
无人控制数据清洗步骤，用于从第二大数据中删除所述对象不具备的动力推进方式及其对应的其他数据，得到第四大数据；
有效历史大数据生成步骤，用于将第三大数据和第四大数据作为有效历史大数据。
优选地，所述对象为车；所述车包括无人车；所述预设类型的与所述对象相关的数据包括所述车当前所在路段的路况数据、所述车当前所在路段的排气污染控制指标数据、所述车当前所在路段的噪音控制指标数据、所述车当前所在路段的限速范围、所述车当前不同动力推进类型所剩余的能量数据、所述车的车型、所述车当前预设数据中的一种或几种。
优选地，所述对象为船；所述船包括无人船；所述预设类型的与所述对象相关的数据包括所述船当前所在航段的海况数据、所述船当前所在航段的排气污染控制指标数据、所述船当前所在航段的噪音控制指标数据、所述船当前所在航段的气象数据、所述船当前所在航段的风力数据、所述船当前需求数据、所述船当前不同动力推进类型所剩余的能量数据、所述船的船型、所述船当前预设数据中的一种或几种。
优选地，所述对象为飞机；所述飞机包括无人机；所述预设类型的与所述对象相关的数据包括所述飞机当前所在航段的天气数据、所述飞机当前所在航段的噪音污染控制指标数据、当前所在航段的风力数据、所述飞机当前不同动力推进类型所剩余的能量数据、机型、所述飞机当前预设数据中的一种或几种。
第二方面，本发明实施例提供一种系统，其特征在于，所述系统执行第一方面任一项所述的混合动力推进方法中的步骤；所述系统包括机器人系统。
本发明实施例具有的优点和有益效果包括：
本发明实施例通过从历史大数据学习得到预设模型，进而通过预设模型和目前数据计算得到目前应该采用的动力推进方式，而且所述历史数据和目前数据中包括动力推进系统所属对象本身的数据和环境数据，从而使得得到的预设模型和推荐的动力推进方式更符合对象和环境的需要、更高效，因此本发明实施例可使得动力推进方式的切换更为智能、高效。例如，对于无人船的混合动力推进的切换而言，需要考虑到的数据包括所述船当前所在航段的海况数据、当前所在航段的排气污染控制指标数据、当前所在航段的噪音控制指标数据、当前所在航段的气象数据、当前所在航段的风力数据、当前其他需求数据、当前不同动力推进类型所剩余的能量数据、船型、当前其他预设数据、等等中的一种或几种。其中，所述航段可以替换为海域。当前其他预设数据包括作战时对噪音、速度等的控制指标数据。而现有技术只能手动切换或者根据速度进行自动切换，手动切换对用户要求高、不智能，而如果仅仅根据速度进行自动切换，只考虑到了速度因素，不适合更广泛的应用场景，在很多应用场景下不但要考虑速度，还要考虑到更多环境的因素和动力推进系统所属对象本身的因素。
本发明实施例提供的基于大数据和人工智能的混合动力推进方法和机器人系统，包括：获取动力推进装置所属的对象的类型，获取预设类型的与所述对象相关的数据，获取预设模型，将所述预设类型的与所述对象相关的数据输入所述预设模型，通过所述预设模型计算得到的输出作为推荐的动力推进方式。上述方法和系统通过基于大数据和人工智能的混合动力推进技术，提高了混合动力推进方式切换的智能性、高效性。
附图说明
图1为本发明的实施例2提供的混合动力推进方法的流程图；
图2为本发明的实施例4提供的预设模型获取步骤的流程图；
图3为本发明的实施例5提供的预设模型获取步骤的流程图；
图4为本发明的实施例5提供的模型计算步骤的流程图；
图5为本发明的实施例11提供的混合动力推进系统的原理框图；
图6为本发明的实施例13提供的预设模型模块的原理框图；
图7为本发明的实施例14提供的预设模型模块的原理框图；
图8为本发明的实施例14提供的模型计算模块的原理框图。
具体实施方式
下面结合本发明实施方式，对本发明实施例中的技术方案进行详细地描述。
(一)本发明的各种实施例中的方法包括以下步骤的各种组合：
实施例1：
一种混合动力推进方法，包括对象类型获取步骤S100、数据获取步骤S200、预设模型获取步骤S300、模型计算步骤S400、系统控制步骤S500。
对象类型获取步骤S100，用于获取动力推进装置所属的对象的类型。所述动力推进装置是所述对象的动力推进装置。所述动力推进装置安装在所述对象上，所以所述对象是动力推进装置所属的对象。所述对象包括车辆、轮船、飞机等等交通工具或者其他需要安装动力推进装置的系统或设备。从而根据对象类型确定所述类型的对象的动力推进装置的动力推进方式，并确定与推进方式相关的预设类型的与所述对象相关的数据。
数据获取步骤S200，用于获取预设类型的与所述对象相关的数据。从而根据预设类型的与所述对象相关的数据计算得到推荐的动力推进方式。
预设模型获取步骤S300，用于获取预设模型。从而通过预设模型在预设类型的与所述对象相关的数据与动力推进方式之间建立对应关系。其中，所述预设模型的输入格式为预设类型的与所述对象相关的数据格式，输出格式为动力推进方式的数据格式；动力推进方式的数据格式可以使用数字格式，将每一个 动力推进方式编码成一个数字；动力推进方式(即动力推进的方式)包括使用电能来推进、柴油机来推进、燃气轮机来推进等等、以及一种或多种推进方式混合的推进方式；
模型计算步骤S400，用于将所述预设类型的与所述对象相关的数据输入所述预设模型，通过所述预设模型计算得到的输出作为推荐的动力推进方式。从而为控制所述动力推进系统提供推荐的动力推进方式。其中，所述输出为所述预设模型的输出；
实施例2：
根据实施例1所述的方法，还包括系统控制步骤S500，如图1所示。
系统控制步骤S500，用于根据所述推荐的动力推进方式控制所述动力推进装置；从而使得所述动力推进系统能以更优的方式运行，从而提高所述对象的所述动力推进系统的智能性和高效性。具体用于判断所述推荐的动力推进方式与当前动力推进方式是否一致：否，则向动力推进装置发送将当前动力推进方式切换为所述推荐的动力推进方式的控制指令。
实施例3：
根据实施例1所述的方法，其中，
所述预设类型由用户预先设置或从知识库中获取；
所述预设类型的数据包括与动力推进方式的选择有相关性的数据；
所述与所述对象相关的数据包括所述对象的数据、所述对象的环境数据；
所述预设类型的与所述对象相关的数据为所述预设类型的与所述对象相关的当前数据或近期数据或最近预设时段内的数据。
实施例4：
根据实施例1所述的方法，
其中，预设模型获取步骤S300包括历史大数据获取步骤S311、对应数据获 取步骤S312、深度学习模型初始化步骤S313、无监督训练步骤S314、有监督训练步骤S315、预设模型生成步骤S316，如图2所示。
历史大数据获取步骤S311，用于获取所述对象所属类型的所有对象的有效历史大数据；
对应数据获取步骤S312，用于获取所述有效历史大数据中预设类型的与所述对象相关的数据及其对应的动力推进方式，其中，所述预设类型的与所述对象相关的数据为预设类型的历史数据；
深度学习模型初始化步骤S313，用于初始化深度学习模型；
无监督训练步骤S314，用于将所述有效历史大数据中所述预设类型的与所述对象相关的数据作为所述深度学习模型的输入，对所述深度学习模型进行无监督训练；
有监督训练步骤S315，用于将所述历史大数据中所述预设类型的与所述对象相关的数据及其对应的动力推进方式分别作为所述深度学习模型的输入和输出，对通过无监督训练之后的所述深度学习模型进行有监督训练；
预设模型生成步骤S316，用于获取有监督训练之后的所述深度学习模型作为所述预设模型。
其中，所述历史大数据可以通过网络在线获取或从历史大数据库获取。所述有效历史大数据包括预设类型的与所述对象相关的数据及其对应的动力推进方式；所述有效历史大数据是在过去很长一段时间内采集的。预设类型的与所述对象相关的数据及其对应的动力推进方式是每一个被采集的对象在每一个被采集的时刻所采集到的，既要采集每一个被采集的对象在每一个被采集的时刻的预设类型的与所述对象相关的数据，还要采集所述每一个被采集的对象在所述每一个被采集的时刻的动力推进方式，其中，所述每一个被采集的对象属于所述对象所属类型的对象集合。
其中，所述对象所属类型包括与所述对象具有至少一种共同的动力推进方式的对象(包括交通工具)。
实施例5：
根据实施例1所述的方法，
其中，预设模型获取步骤S300包括历史大数据获取步骤S321、对应数据获取步骤S322、模型数据设置步骤S323，如图3所示。
历史大数据获取步骤S321，用于获取所述对象所属类型的所有对象的有效历史大数据；
对应数据获取步骤S322，用于获取所述有效历史大数据中预设类型的与所述对象相关的数据及其对应的动力推进方式；其中，所述预设类型的与所述对象相关的数据为预设类型的历史数据；
模型数据设置步骤S323，用于将所述有效历史大数据中所述预设类型的与所述对象相关的数据及其对应的动力推进方式分别作为所述预设模型的待匹配数据及其对应的待推荐数据。
其中，模型计算步骤S400包括匹配步骤S421、选取步骤S422、推荐步骤S423，如图4所示。
匹配步骤S421，用于所述将所述预设类型的与所述对象相关的数据与所述预设模型中的每一个所述待匹配数据进行模糊匹配；
选取步骤S422，用于选取与所述输入的所述预设类型的与所述对象相关的数据匹配度最大的所述预设模型中的所述待匹配数据；
推荐步骤S423，用于从所述预设模型中获取与所选取的所述待匹配数据所对应的待推荐数据作为所述预设模型计算得到的输出，将所述输出作为推荐的动力推进方式。
实施例4采用的是大数据、深度学习技术，实施例5采用的是大数据及其推荐技术。
实施例6：
根据实施例4或5所述的方法，
历史大数据获取步骤S321包括：
有人控制数据获取步骤S321-1：用于获取“有人控制(例如有人驾驶)”且“与所述对象具有至少一种共同的动力推进方式”的所有同类对象(例如车)的历史数据作为第一大数据；所述同类对象为所述对象所属类型的对象；所述同类对象与所述对象都属于所述对象所属类型；所述历史数据包括所述预设类型的数据及其对应的动力推进方式；因为人是有智能的，对象在有人控制(例如有人驾驶)时所选择的动力推进方式更可信，而无人机、无人车、无人船等无人控制(例如无人驾驶)的对象(例如车)的历史大数据不一定可信。所述有人控制包括有人进行动力推进方式的选择；
无人控制数据获取步骤S321-2：获取“无人控制(例如无人驾驶)”且“动力推进方式的选择效果满足预设条件”且“与所述对象具有至少一种共同的动力推进方式”的所有同类对象(例如车)的历史数据作为第二大数据；所述历史数据包括所述预设类型的数据及其对应的动力推进方式；其中，动力推进方式的选择效果满足预设条件包括动力推进方式的选择效果的用户评分大于预设阈值。所述无人控制包括无人进行动力推进方式的选择；
有人控制数据清洗步骤S321-3：从第一大数据中删除所述对象不具备的动力推进方式及其对应的其他数据，包括从第一大数据中删除所述对象不具备的动力推进方式及其对应的所述预设类型的数据，得到第三大数据；
无人控制数据清洗步骤S321-4：从第二大数据中删除所述对象不具备的动力推进方式及其对应的其他数据，包括从第二大数据中删除所述对象不具备的动力推进方式及其对应的所述预设类型的数据，得到第四大数据；
有效历史大数据生成步骤S321-5：将第三大数据和第四大数据作为有效历史大数据。
实施例7：
根据实施例1所述的方法，
其中，所述对象为车；所述车包括无人车；
其中，所述预设类型的与所述对象相关的数据包括所述车当前所在路段的 路况数据、当前所在路段的排气污染控制指标数据、当前所在路段的噪音控制指标数据、当前所在路段的限速范围、当前不同动力推进类型所剩余的能量数据、车型、当前其他预设数据、等等中的一种或几种。其中，所述道段可以替换为区域。当前其他预设数据包括作战时对噪音、速度等的控制指标数据。
实施例8：
根据实施例1所述的方法，
其中，所述对象为船；所述船包括无人船；
其中，所述预设类型的与所述对象相关的数据包括所述船当前所在航段的海况数据、当前所在航段的排气污染控制指标数据、当前所在航段的噪音控制指标数据、当前所在航段的气象数据、当前所在航段的风力数据、当前其他需求数据、当前不同动力推进类型所剩余的能量数据、船型、当前其他预设数据、等等中的一种或几种。其中，所述航段可以替换为海域。当前其他预设数据包括作战时对噪音、速度等的控制指标数据。
实施例9：
根据实施例1所述的方法，
其中，所述对象为飞机；所述飞机包括无人机；
其中，所述预设类型的与所述对象相关的数据包括所述飞机当前所在航段的天气数据、当前所在航段的噪音污染控制指标数据、当前所在航段的风力数据、当前不同动力推进类型所剩余的能量数据、机型、当前其他预设数据、等等中的一种或几种。其中，所述航段可以替换为空域。当前其他预设数据包括作战时对噪音、速度等的控制指标数据。
实施例10：
一种混合动力推进系统，包括对象类型获取模块100、数据获取模块200、预设模型获取模块300、模型计算模块400。
对象类型获取模块100，用于获取动力推进装置所属的对象的类型。
数据获取模块200，用于获取预设类型的与所述对象相关的数据。
预设模型获取模块300，用于获取预设模型。
模型计算模块400，用于将所述预设类型的与所述对象相关的数据输入所述预设模型，通过所述预设模型计算得到的输出作为推荐的动力推进方式。
实施例11：
根据实施例10所述的系统，还包括系统控制模块500，如图5所示。
系统控制模块500，用于根据所述推荐的动力推进方式控制所述动力推进装置。具体为，判断所述推荐的动力推进方式与当前动力推进方式是否一致：否，则向动力推进装置发送将当前动力推进方式切换为所述推荐的动力推进方式的控制指令。
实施例12：
根据实施例10所述的系统，其中，
所述预设类型由用户预先设置或从知识库中获取；
所述预设类型的数据包括与动力推进方式的选择有相关性的数据；
所述与所述对象相关的数据包括所述对象的数据、所述对象的环境数据；
所述预设类型的与所述对象相关的数据为所述预设类型的与所述对象相关的当前数据或近期数据或最近预设时段内的数据。
实施例13：
根据实施例10所述的系统，
其中，预设模型获取模块300包括历史大数据获取模块311、对应数据获取模块312、深度学习模型初始化模块313、无监督训练模块314、有监督训练模块315、预设模型生成模块316，如图6所示。
历史大数据获取模块311，用于获取所述对象所属类型的所有对象的有效历 史大数据；
对应数据获取模块312，用于获取所述有效历史大数据中预设类型的与所述对象相关的数据及其对应的动力推进方式；
深度学习模型初始化模块313，用于初始化深度学习模型；
无监督训练模块314，用于将所述有效历史大数据中所述预设类型的与所述对象相关的数据作为所述深度学习模型的输入，对所述深度学习模型进行无监督训练；
有监督训练模块315，用于将所述历史大数据中所述预设类型的与所述对象相关的数据及其对应的动力推进方式分别作为所述深度学习模型的输入和输出，对通过无监督训练之后的所述深度学习模型进行有监督训练；
预设模型生成模块316，用于获取有监督训练之后的所述深度学习模型作为所述预设模型。
实施例14：
根据实施例10所述的系统，
其中，预设模型获取模块300包括历史大数据获取模块321、对应数据获取模块322、模型数据设置模块323，如图7所示。
历史大数据获取模块321，用于获取所述对象所属类型的所有对象的有效历史大数据；
对应数据获取模块322，用于获取所述有效历史大数据中预设类型的与所述对象相关的数据及其对应的动力推进方式；
模型数据设置模块323，用于将所述有效历史大数据中所述预设类型的与所述对象相关的数据及其对应的动力推进方式分别作为所述预设模型的待匹配数据及其对应的待推荐数据。
其中，模型计算模块400包括匹配模块421、选取模块422、推荐模块423，如图8所示。
匹配模块421，用于所述将所述预设类型的与所述对象相关的数据与所述预 设模型中的每一个所述待匹配数据进行模糊匹配；
选取模块422，用于选取与所述输入的所述预设类型的与所述对象相关的数据匹配度最大的所述预设模型中的所述待匹配数据；
推荐模块423，用于从所述预设模型中获取与所选取的所述待匹配数据所对应的待推荐数据作为所述预设模型计算得到的输出，将所述输出作为推荐的动力推进方式。
实施例15：
根据实施例13或14所述的系统，
其中，历史大数据获取模块321包括：
有人控制数据获取模块321-1：用于获取“有人控制(例如有人驾驶)”且“与所述对象具有至少一种共同的动力推进方式”的所有同类对象(例如车)的历史数据作为第一大数据；所述同类对象为所述对象所属类型的对象；所述同类对象与所述对象都属于所述对象所属类型；所述历史数据包括所述预设类型的数据及其对应的动力推进方式；
无人控制数据获取模块321-2：获取“无人控制(例如无人驾驶)”且“动力推进方式的选择效果满足预设条件”且“与所述对象具有至少一种共同的动力推进方式”的所有同类对象(例如车)的历史数据作为第二大数据；所述历史数据包括所述预设类型的数据及其对应的动力推进方式；
有人控制数据清洗模块321-3：从第一大数据中删除所述对象不具备的动力推进方式及其对应的其他数据，包括从第一大数据中删除所述对象不具备的动力推进方式及其对应的所述预设类型的数据，得到第三大数据；
无人控制数据清洗模块321-4：从第二大数据中删除所述对象不具备的动力推进方式及其对应的其他数据，包括从第二大数据中删除所述对象不具备的动力推进方式及其对应的所述预设类型的数据，得到第四大数据；
有效历史大数据生成模块321-5：将第三大数据和第四大数据作为有效历史大数据。
实施例16：
根据实施例10所述的系统，包括：
其中，所述对象为车；所述车包括无人车；
其中，所述预设类型的与所述对象相关的数据包括所述车当前所在路段的路况数据、当前所在路段的排气污染控制指标数据、当前所在路段的噪音控制指标数据、当前所在路段的限速范围、当前不同动力推进类型所剩余的能量数据、车型、当前其他预设数据、等等中的一种或几种。其中，所述道段可以替换为区域。当前其他预设数据包括作战时对噪音、速度等的控制指标数据。
实施例17：
根据实施例10所述的系统，
其中，所述对象为船；所述船包括无人船；
其中，所述预设类型的与所述对象相关的数据包括所述船当前所在航段的海况数据、当前所在航段的排气污染控制指标数据、当前所在航段的噪音控制指标数据、当前所在航段的气象数据、当前所在航段的风力数据、当前其他需求数据、当前不同动力推进类型所剩余的能量数据、船型、当前其他预设数据、等等中的一种或几种。其中，所述航段可以替换为海域。当前其他预设数据包括作战时对噪音、速度等的控制指标数据。
实施例18：
根据实施例10所述的系统，
其中，所述对象为飞机；所述飞机包括无人机；
其中，所述预设类型的与所述对象相关的数据包括所述飞机当前所在航段的天气数据、当前所在航段的噪音污染控制指标数据、当前所在航段的风力数据、当前不同动力推进类型所剩余的能量数据、机型、当前其他预设数据、等等中的一种或几种。其中，所述航段可以替换为空域。当前其他预设数据包括作战时对噪音、速度等的控制指标数据。
实施例19：
提供一种机器人系统，所述机器人中分别配置有如实施例10至实施例18所述的系统。
上述各实施例中的方法和系统可以在计算机、服务器、云服务器、超级计算机、机器人、嵌入式设备、电子设备等上执行和部署。
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本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 (10)

  1. 一种混合动力推进方法，其特征在于，所述方法包括：
    对象类型获取步骤，用于获取动力推进装置所属的对象的类型；
    数据获取步骤，用于获取预设类型的与所述对象相关的数据；
    预设模型获取步骤，用于获取预设模型；
    模型计算步骤，用于将所述预设类型的与所述对象相关的数据输入所述预设模型，通过所述预设模型计算得到的输出作为推荐的动力推进方式。
  2. 根据权利要求1所述的混合动力推进方法，其特征在于，所述方法还包括：
    系统控制步骤，用于根据所述推荐的动力推进方式控制所述动力推进装置。
  3. 根据权利要求1所述的混合动力推进方法，其特征在于，
    所述预设类型由用户预先设置或从知识库中获取；
    所述预设类型的数据包括与动力推进方式的选择有相关性的数据；
    所述与所述对象相关的数据包括所述对象的数据、所述对象的环境数据；
    所述预设类型的与所述对象相关的数据为所述预设类型的与所述对象相关的当前数据或近期数据或最近预设时段内的数据。
  4. 根据权利要求1所述的混合动力推进方法，其特征在于，所述预设模型获取步骤包括：
    历史大数据获取步骤，用于获取所述对象所属类型的所有对象的有效历史大数据；
    对应数据获取步骤，用于获取所述有效历史大数据中预设类型的数据及其对应的动力推进方式；
    深度学习模型初始化步骤，用于初始化深度学习模型；
    无监督训练步骤，用于将所述有效历史大数据中所述预设类型的数据作为所述深度学习模型的输入，对所述深度学习模型进行无监督训练；
    有监督训练步骤，用于将所述历史大数据中所述预设类型的数据及其对应的动力推进方式分别作为所述深度学习模型的输入和输出，对通过无监督训练之后的所述深度学习模型进行有监督训练；
    预设模型生成步骤，用于获取有监督训练之后的所述深度学习模型作为所述预设模型。
  5. 根据权利要求1所述的混合动力推进方法，其特征在于，
    所述预设模型获取步骤包括：
    历史大数据获取步骤，用于获取所述对象所属类型的所有对象的有效历史大数据；
    对应数据获取步骤，用于获取所述有效历史大数据中预设类型的数据及其对应的动力推进方式；
    模型数据设置步骤，用于将所述有效历史大数据中所述预设类型的数据及其对应的动力推进方式分别作为所述预设模型的待匹配数据及其对应的待推荐数据；
    所述模型计算步骤包括：
    匹配步骤，用于所述将所述预设类型的与所述对象相关的数据与所述预设模型中的每一个所述待匹配数据进行模糊匹配；
    选取步骤，用于选取与所述输入的所述预设类型的与所述对象相关的数据匹配度最大的所述预设模型中的所述待匹配数据；
    推荐步骤，用于从所述预设模型中获取与所选取的所述待匹配数据所对应的待推荐数据作为所述预设模型计算得到的输出，将所述输出作为推荐的动力推进方式。
  6. 根据权利要求4或5所述的混合动力推进方法，其特征在于，历史大数据获取步骤包括：
    有人控制数据获取步骤，用于获取“有人控制”且“与所述对象具有至少一种共同的动力推进方式”的所有同类对象的历史数据作为第一大数据；
    无人控制数据获取步骤，用于获取“无人控制”且“动力推进方式的选择效果满足预设条件”且“与所述对象具有至少一种共同的动力推进方式”的所有同类对象的历史数据作为第二大数据；
    有人控制数据清洗步骤，用于从第一大数据中删除所述对象不具备的动力 推进方式及其对应的其他数据，得到第三大数据；
    无人控制数据清洗步骤，用于从第二大数据中删除所述对象不具备的动力推进方式及其对应的其他数据，得到第四大数据；
    有效历史大数据生成步骤，用于将第三大数据和第四大数据作为有效历史大数据。
  7. 根据权利要求1所述的混合动力推进方法，其特征在于，所述对象为车；所述车包括无人车；所述预设类型的与所述对象相关的数据包括所述车当前所在路段的路况数据、所述车当前所在路段的排气污染控制指标数据、所述车当前所在路段的噪音控制指标数据、所述车当前所在路段的限速范围、所述车当前不同动力推进类型所剩余的能量数据、所述车的车型、所述车当前预设数据中的一种或几种。
  8. 根据权利要求1所述的混合动力推进方法，其特征在于，所述对象为船；所述船包括无人船；所述预设类型的与所述对象相关的数据包括所述船当前所在航段的海况数据、所述船当前所在航段的排气污染控制指标数据、所述船当前所在航段的噪音控制指标数据、所述船当前所在航段的气象数据、所述船当前所在航段的风力数据、所述船当前需求数据、所述船当前不同动力推进类型所剩余的能量数据、所述船的船型、所述船当前预设数据中的一种或几种。
  9. 根据权利要求1所述的混合动力推进方法，其特征在于，所述对象为飞机；所述飞机包括无人机；所述预设类型的与所述对象相关的数据包括所述飞机当前所在航段的天气数据、所述飞机当前所在航段的噪音污染控制指标数据、当前所在航段的风力数据、所述飞机当前不同动力推进类型所剩余的能量数据、机型、所述飞机当前预设数据中的一种或几种。
  10. 一种系统，其特征在于，所述系统执行权利要求1-9任一项所述的混合动力推进方法中的步骤；所述系统包括机器人系统。
PCT/CN2020/097978 2019-08-29 2020-06-24 基于大数据和人工智能的混合动力推进方法和机器人系统 WO2021036456A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910794455.8A CN110509913B (zh) 2019-08-29 2019-08-29 基于大数据和人工智能的混合动力推进方法和机器人系统
CN201910794455.8 2019-08-29

Publications (1)

Publication Number Publication Date
WO2021036456A1 true WO2021036456A1 (zh) 2021-03-04

Family

ID=68628065

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/097978 WO2021036456A1 (zh) 2019-08-29 2020-06-24 基于大数据和人工智能的混合动力推进方法和机器人系统

Country Status (2)

Country Link
CN (1) CN110509913B (zh)
WO (1) WO2021036456A1 (zh)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110509913B (zh) * 2019-08-29 2021-05-21 南京智慧光信息科技研究院有限公司 基于大数据和人工智能的混合动力推进方法和机器人系统
CN110737260B (zh) * 2019-08-29 2022-02-11 南京智慧光信息科技研究院有限公司 基于大数据和人工智能的自动作业方法和机器人系统

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104554251A (zh) * 2014-12-09 2015-04-29 河南理工大学 基于道路坡度信息的混合动力汽车节能预测控制方法
US20170320481A1 (en) * 2014-11-06 2017-11-09 Volvo Truck Corporation A hybrid vehicle and a method for energy management of a hybrid vehicle
CN107862864A (zh) * 2017-10-18 2018-03-30 南京航空航天大学 基于驾驶习惯和交通路况的行驶工况智能预测估计方法
CN108108841A (zh) * 2017-12-19 2018-06-01 天津大学 一种基于大数据库的混合动力能量管理策略全局优化系统
CN108177648A (zh) * 2018-01-02 2018-06-19 北京理工大学 一种基于智能预测的插电式混合动力车辆的能量管理方法
CN108248609A (zh) * 2016-12-29 2018-07-06 现代自动车株式会社 混合动力车辆和在混合动力车辆中预测驾驶样式的方法
CN110509913A (zh) * 2019-08-29 2019-11-29 南京智慧光信息科技研究院有限公司 基于大数据和人工智能的混合动力推进方法和机器人系统

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104364668B (zh) * 2012-06-13 2017-02-22 株式会社Lg化学 估计包括混合正极材料的二次电池的电压的设备和方法
US20150161742A1 (en) * 2013-12-06 2015-06-11 Mastercard International Incorporated Automatic determination of vehicle information based on transaction information
US9878632B2 (en) * 2014-08-19 2018-01-30 General Electric Company Vehicle propulsion system having an energy storage system and optimized method of controlling operation thereof
DE102014222513B4 (de) * 2014-11-04 2020-02-20 Continental Automotive Gmbh Verfahren zum Betrieb eines Hybrid- oder Elektrofahrzeugs
US10394243B1 (en) * 2018-09-21 2019-08-27 Luminar Technologies, Inc. Autonomous vehicle technology for facilitating operation according to motion primitives

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170320481A1 (en) * 2014-11-06 2017-11-09 Volvo Truck Corporation A hybrid vehicle and a method for energy management of a hybrid vehicle
CN104554251A (zh) * 2014-12-09 2015-04-29 河南理工大学 基于道路坡度信息的混合动力汽车节能预测控制方法
CN108248609A (zh) * 2016-12-29 2018-07-06 现代自动车株式会社 混合动力车辆和在混合动力车辆中预测驾驶样式的方法
CN107862864A (zh) * 2017-10-18 2018-03-30 南京航空航天大学 基于驾驶习惯和交通路况的行驶工况智能预测估计方法
CN108108841A (zh) * 2017-12-19 2018-06-01 天津大学 一种基于大数据库的混合动力能量管理策略全局优化系统
CN108177648A (zh) * 2018-01-02 2018-06-19 北京理工大学 一种基于智能预测的插电式混合动力车辆的能量管理方法
CN110509913A (zh) * 2019-08-29 2019-11-29 南京智慧光信息科技研究院有限公司 基于大数据和人工智能的混合动力推进方法和机器人系统

Also Published As

Publication number Publication date
CN110509913A (zh) 2019-11-29
CN110509913B (zh) 2021-05-21

Similar Documents

Publication Publication Date Title
WO2021036456A1 (zh) 基于大数据和人工智能的混合动力推进方法和机器人系统
CN110737986B (zh) 一种无人船能效智能优化仿真系统及方法
US20220073052A1 (en) Artificial intelligence amphibious vertical take-off and landing modular hybrid flying automobile
CN111025969B (zh) 一种基于信息融合的野生动物监测系统及方法
KR102438842B1 (ko) 하이브리드 선박
WO2021036543A1 (zh) 基于大数据和人工智能的自动作业方法和机器人系统
CN107229223B (zh) 一种海洋能无人艇多驱动模式的智能切换系统
CN106647809B (zh) 一种基于任务规划的无人机用混合动力系统能量优化方法
CN112611080A (zh) 一种基于联邦学习的智能空调控制系统和方法
CN114545974B (zh) 一种基于智能机场的有能力约束无人机弧路径规划方法
CN110502015B (zh) 基于大数据和人工智能的速度控制方法和机器人系统
CN106451722A (zh) 一种无人机电池管理系统及管理方法
CN113872198A (zh) 一种基于强化学习方法的主动配电网故障恢复方法
CN108394429B (zh) 一种为城轨列车群生成自动驾驶曲线的方法
CN110488828B (zh) 基于大数据和人工智能的航行灯控制方法和机器人系统
CN114882759A (zh) 虚实混合一体化仿真的智能船多通道交互模拟系统和方法
CN110488803B (zh) 基于大数据和人工智能的制动方法和机器人系统
CN110488804B (zh) 基于大数据和人工智能的联合导航方法和机器人系统
CN111208729B (zh) 一种绝缘斗温控装置自适应控制方法及装置
TW201544399A (zh) 多電源無人飛機
CN114898594B (zh) 一种可搭载无人机的通感算控一体化智能灯船控制系统
Kagabo et al. Trajectory determination for energy efficient autonomous soaring
CN112821456B (zh) 基于迁移学习的分布式源-储-荷匹配方法及装置
CN109816934A (zh) 一种防驾驶疲劳的预警装置及预警方法
CN115934691A (zh) 短期光伏功率的确定方法及装置

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: 20856982

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: 20856982

Country of ref document: EP

Kind code of ref document: A1

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 08/08/2022)

122 Ep: pct application non-entry in european phase

Ref document number: 20856982

Country of ref document: EP

Kind code of ref document: A1