CN108803328A - Camera self-adapting regulation method, device and camera - Google Patents
Camera self-adapting regulation method, device and camera Download PDFInfo
- Publication number
- CN108803328A CN108803328A CN201810614204.2A CN201810614204A CN108803328A CN 108803328 A CN108803328 A CN 108803328A CN 201810614204 A CN201810614204 A CN 201810614204A CN 108803328 A CN108803328 A CN 108803328A
- Authority
- CN
- China
- Prior art keywords
- intensified learning
- camera
- parameter
- state information
- regulation method
- Prior art date
- Legal status (The legal status 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 status listed.)
- Granted
Links
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
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/04—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators
- G05B13/042—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators in which a parameter or coefficient is automatically adjusted to optimise the performance
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Studio Devices (AREA)
Abstract
A kind of camera self-adapting regulation method of present invention offer, device, camera and computer storage media, the camera self-adapting regulation method include：The intensified learning model that the adjustment action input of the kernel state information of multiple determinations and corresponding initial policy is pre-established generates intensified learning parameter；When adjust automatically, using current kernel state information, the intensified learning model and the intensified learning parameter, optimal correction action is obtained.The camera self-adapting regulation method of the present invention allows camera by intensified learning model and intensified learning parameter, autonomous to judge that the trend of environmental change makes corresponding adjustment, the abnormal probability with failure of reduction.
Description
Technical field
The present invention relates to intelligent security guard technical fields, in particular to a kind of camera self-adapting regulation method, dress
It sets, camera and computer storage media.
Background technology
With the continuous development of security and guard technology and universal, monitoring camera has spread all over each corner in city.
But the monitoring camera positioned at security protection front end is often link the weakest in entire video monitoring system, camera is easy
It is influenced by weather, environment, once there is exception or failure, the safety monitoring of relevant range is as illusory.
In existing technology, various ancillary hardwares, example can be arranged in the influence for weather and environment in camera
It is such as provided with fan in camera, for radiating, weather overheat is avoided to cause the exception or failure of camera.But it images
Algorithm that various ancillary hardwares work or application program are controlled in head and is all based on simple strategy, for example, only in temperature
When a certain predetermined temperature, which just enters the state of normal work.This adjustable strategies are too simple, no
The scene of various camera applications can be met, video camera cannot make corresponding adjustment according to the trend of environmental change, can not
The abnormal probability with failure of camera is reduced well.
Invention content
In view of the above problems, the present invention provides a kind of camera self-adapting regulation method, device, camera and computers
Storage medium reduces abnormal and failure probability so that video camera independently judges that the trend of environmental change makes corresponding adjustment.
To achieve the goals above, the present invention adopts the following technical scheme that：
A kind of camera self-adapting regulation method, including：
The adjustment action input of the kernel state information of multiple determinations and corresponding initial policy is pre-established strong
Change learning model, generates intensified learning parameter；
When adjust automatically, joined using current kernel state information, the intensified learning model and the intensified learning
Number obtains optimal correction action.
Preferably, the intensified learning model is on-policy linear sarsa intensified learning models.
Preferably, the action value function that the intensified learning model is fitted is：
Above formula S is the kernel state information, and A acts for the adjustment, and w is the intensified learning parameter, and x is study sample
This, i.e., Strengthen the gradient of parameter w.
Preferably, the object function of the intensified learning parameter w of the intensified learning model is：
Above formula qπ(S, A) is the true action value function based on the initial policy, using based on sarsa (λ) algorithm
'sIt estimates：
Above formula Rt+nThe gain brought is acted for adjustment described in the t+n moment, λ is the preset hyper parameter of sarsa (λ) algorithm,
γ is the preset hyper parameter of intensified learning, Q (St+n) beπ is the initial policy.
Preferably, the camera self-adapting regulation method further includes：
According to intensified learning ginseng described in optimal correction action and the current kernel state information updating and optimizing
Number.
Preferably, the formula of the update optimization intensified learning parameter is：
Above formulaTo strengthen the gradient of parameter w.
The present invention also provides a kind of camera self-adapting adjusting apparatus, including：
Parameter generation module, for acting the adjustment of the kernel state information of multiple determinations and corresponding initial policy
The intensified learning model pre-established is inputted, intensified learning parameter is generated；
Automatic regulating module, be used for adjust automatically when, using current kernel state information, the intensified learning model with
And the intensified learning parameter, obtain optimal correction action.
Preferably, the camera self-adapting adjusting apparatus further includes：
Parameter update module, for excellent according to optimal correction action and the current kernel state information update
Change the intensified learning parameter.
The present invention also provides a kind of camera, including memory and processor, the memory is for storing computer
Program, the processor runs the computer program so that the camera executes the camera adaptively side of adjustment
Method.
The present invention also provides a kind of computer storage media, the computer journey that is stored with used in the camera
Sequence.
The present invention provides a kind of camera self-adapting regulation method, this method includes：By the kernel state of multiple determinations
The intensified learning model that the adjustment action input of information and corresponding initial policy pre-establishes generates intensified learning parameter；
When adjust automatically, using current kernel state information, the intensified learning model and the intensified learning parameter, obtain most
Excellent adjustment action.The camera self-adapting regulation method of the present invention, allows camera to pass through intensified learning model and reinforcing
Learning parameter, it is autonomous to judge that the trend of environmental change makes corresponding adjustment, reduce abnormal and failure probability.
To enable the above objects, features and advantages of the present invention to be clearer and more comprehensible, preferred embodiment cited below particularly, and coordinate
Appended attached drawing, is described in detail below.
Description of the drawings
In order to illustrate the technical solution of the embodiments of the present invention more clearly, below will be to needed in the embodiment attached
Figure is briefly described, it should be understood that the following drawings illustrates only certain embodiments of the present invention, therefore is not construed as pair
The restriction of the scope of the invention.
Fig. 1 is a kind of camera provided in an embodiment of the present invention；
Fig. 2 is a kind of flow chart for camera self-adapting regulation method that the embodiment of the present invention 1 provides；
Fig. 3 is a kind of flow chart for camera self-adapting regulation method that the embodiment of the present invention 2 provides；
Fig. 4 is a kind of structure chart for camera self-adapting adjusting apparatus that the embodiment of the present invention 3 provides；
Fig. 5 is the structure chart for another camera self-adapting adjusting apparatus that the embodiment of the present invention 3 provides.
Specific implementation mode
Below in conjunction with attached drawing in the embodiment of the present invention, technical solution in the embodiment of the present invention carries out clear, complete
Ground describes, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.Usually exist
The component of the embodiment of the present invention described and illustrated in attached drawing can be arranged and be designed with a variety of different configurations herein.Cause
This, the detailed description of the embodiment of the present invention to providing in the accompanying drawings is not intended to limit claimed invention below
Range, but it is merely representative of the selected embodiment of the present invention.Based on the embodiment of the present invention, those skilled in the art are not doing
The every other embodiment obtained under the premise of going out creative work, shall fall within the protection scope of the present invention.
Following each embodiments can be applied in camera as shown in Figure 1, and Fig. 1 shows the structural frames of the camera
Figure, the camera 100 include：Ethernet interface 110, memory 120, sensor 130, voicefrequency circuit 140, Wireless Fidelity
The components such as (wireless fidelity, WiFi) module 150, processor 160 and power supply 170.Those skilled in the art can
To understand, 100 structure of camera shown in Fig. 1 does not constitute the restriction to camera, may include more or more than illustrating
Few component either combines certain components or different components arrangement.
Embodiment 1
Fig. 2 is a kind of flow chart for camera self-adapting regulation method that the embodiment of the present invention 1 provides, and this method includes such as
Lower step：
Step S21：The adjustment action input of the kernel state information of multiple determinations and corresponding initial policy is advance
The intensified learning model of foundation generates intensified learning parameter.
In the embodiment of the present invention, the kernel state information of camera includes temperature, humidity, fan operating state and heating
Device working condition etc..Wherein, the temperature information of acquisition includes the DIE Temperature information of camera and the temperature of each key component
Spend information；The humidity information of acquisition includes the core humidity information of camera and the humidity information of each key component；It obtains
Fan operating state include fan work operation power and gear, which can be divided into 5 grades according to power；What is obtained adds
Hot device working condition includes the power and gear of heater work operation, which can be divided into 5 grades according to power.
The kernel state information of the camera can be acquired using sensor, it is, for example, possible to use multiple humidity sensors
The core humidity of device acquisition camera and the humidity of each key component, and multiple humidity thresholds can be set, work as humidity
When in one of humidity threshold, then it can start the adjustment that corresponding dehumidifying strategy carries out heater and fan.Its
In, algorithm or application program can be used to detect various kernel state information in real time in above-mentioned camera, for example, using journey is applied
Sequence handles the information of each sensor in real time, monitors various kernel state information in real time, and it is possible to every time carry out heater or
When the adjustment of fan, each kernel state information after record adjustment action, each kernel state information of adjustment preceding camera and adjustment,
Training sample as intensified learning is stored.
In the embodiment of the present invention, which refers to original in camera by maintenance staff or camera being gone out
The preset countermeasure in face of camera exception kernel state before factory, for example, when temperature is more than a scheduled temperature value,
Increase fan power or gear, reduces heater power or gear；When temperature is less than a scheduled temperature value, fan is reduced
Power or gear increase heater power or gear；When humidity is more than a scheduled humidity value, increase fan power or shelves
Position increases heater power or gear.
In the embodiment of the present invention, intensified learning model be on-policy linear sarsa (on-policy linear,
It is tactful on linear line) intensified learning model.It is being updated using the intensified learning model of on-policy linear sarsa algorithms
The status information after current adjustment action and adjustment can be continued to use when learning parameter as sample, to accomplish camera
Self-teaching makes camera adaptively obtain and is acted in face of the optimal correction of current state information.
The action value function that intensified learning model is fitted is：
Above formula S is the kernel state information, and A acts for the adjustment, and w is the intensified learning parameter, and x is study sample
Originally it is To strengthen the gradient of parameter w.
The action value function of above-mentioned fitting, also as a kind of trend function, for fit under different core state for
The adjustment action that camera should be made.Camera is made correctly by the intensified learning constantly improve function using the function
Adjustment, the process for improving function is the process of intensified learning, and the sample of study is the last sample generated from main modulation
This.In the embodiment of the present invention, an intensified learning parameter w defined in the action value function of fitting, after inputting learning sample
The intensified learning parameter w for generating concrete numerical value can be by continuous using the optimization of the parameter as the main target of intensified learning
Autonomous learning updates intensified learning parameter w, and kernel state information S, adjustment act the letter between A and intensified learning parameter w
Number relationship can use on-policy linear sarsa algorithms to build.
Adjustment action A include fan power increase or decrease one grade, heater power increase or decrease one grade, fan power
Increase or decrease two grades, heater power increase or decrease two grades etc., and various adjustment action A can pass through and quantify conversion
For numerical value, to carry out intensified learning, such as binary number or hexadecimal number can be quantified as etc..
The object function of the intensified learning parameter w of intensified learning model is：
Above formula qπ(S, A) is the true action value function based on the initial policy, using based on sarsa (λ) algorithm
'sIt estimates：
Above formula Rt+nThe gain brought is acted for adjustment described in the t+n moment, λ is the preset hyper parameter of sarsa (λ) algorithm,
γ is the preset hyper parameter of intensified learning, Q (St+n) beπ is the initial policy.N is natural number, such as 1,
2,3 etc..
In the embodiment of the present invention, for the degree of optimization of intensified learning parameter, an object function can be established to carry out
Differentiate, the gain that adjustment action is brought in above formula, actually the differentiation of the kernel state for camera after adjustment action, for example,
When the temperature of a yield value defined in intensified learning model, adjustment action rear camera is in suitable temperature value,
Then the yield value is 1, is otherwise -1.
Hyper parameter λ and hyper parameter γ are configured by maintenance personnel, and optimal hyper parameter can be arranged in maintenance personnel,
To improve the performance and effect of intensified learning model, to improve the autonomous adjustment capability of camera.
When being analyzed for object function J (w), local derviation can be carried out to intensified learning parameter w, be denoted as Δ w：
Step S22：When adjust automatically, joined using current kernel state information, intensified learning model and intensified learning
Number obtains optimal correction action.
In the embodiment of the present invention, camera can be by each kernel state information input intensified learning mould of current time acquisition
In type, optimal correction action is obtained by action value function after intensified learning and intensified learning parameter w.Also, it takes the photograph
As head execute optimal correction action after can acquire each kernel state information, with obtain execute the adjustment action gain, deposit
Kernel state information before storage adjustment action message, adjustment action and corresponding gain, as intensified learning sample.Wherein, should
First intensified learning sample of camera can be the sample that camera executes that initial policy generates, and carry out intensified learning
Afterwards, which substitutes the initial policy and generates optimal correction action and corresponding learning sample, makes intensified learning mould
The continuous adaptive learning of type.
Also, the high-quality learning sample that the intensified learning model can also take maintenance staff to input learns,
With faster more Accurate Curve-fitting, optimal adaptive adjustment is made.
Embodiment 2
Fig. 3 is a kind of flow chart for camera self-adapting regulation method that the embodiment of the present invention 2 provides, and this method includes such as
Lower step：
Step S31：The adjustment action input of the kernel state information of multiple determinations and corresponding initial policy is advance
The intensified learning model of foundation generates intensified learning parameter.
This step is consistent with above-mentioned steps S21, and details are not described herein.
Step S32：When adjust automatically, joined using current kernel state information, intensified learning model and intensified learning
Number obtains optimal correction action.
This step is consistent with above-mentioned steps S22, and details are not described herein.
Step S33：According to optimal correction action and current kernel state information updating and optimizing intensified learning parameter.
In the embodiment of the present invention, camera can all use each biography after executing optimal correction action every time after a period of time
Sensor acquires each kernel state information again, and whether reaches expected after can also judging adjustment by the kernel state information
Effect generates corresponding gain, and it is learning sample to store the kernel state information before optimal correction action and adjustment.
The formula of above-mentioned update optimization intensified learning parameter w is：
WhereinFormula can be passed through Join to strengthen
The gradient of number w, and corresponding learning sample acquire, thus intensified learning parameter w can by learning to be updated every time,
Ensure that camera can be transferred through intensified learning and adaptively adjusted under various circumstances.
Embodiment 3
Fig. 4 is a kind of structure chart for camera self-adapting adjusting apparatus that the embodiment of the present invention 3 provides.
The camera self-adapting adjusting apparatus 400 includes：
Parameter generation module 410 is used for the adjustment of the kernel state information of multiple determinations and corresponding initial policy
The intensified learning model that action input pre-establishes generates intensified learning parameter；
Automatic regulating module 420, be used for adjust automatically when, using current kernel state information, intensified learning model with
And intensified learning parameter, obtain optimal correction action.
As shown in figure 5, the camera self-adapting adjusting apparatus 400 further includes：
Parameter update module 430, for strong according to optimal correction action and current kernel state information updating and optimizing
Change learning parameter.
More specifically explanation can refer to corresponding part in previous embodiment to above-mentioned each module in the present embodiment, herein
It repeats no more.
In addition, the present invention also provides a kind of camera, which includes memory and processor, and memory can be used for
Computer program is stored, processor is by running the computer program, to make camera execute the above method or above-mentioned
The function of modules in camera self-adapting adjusting apparatus.
Memory may include storing program area and storage data field, wherein storing program area can storage program area, at least
Application program (such as sound recording function, picture recording function etc.) needed for one function etc.；Storage data field can store root
Created data (such as audio data etc.) etc. are used according to camera.In addition, memory may include high random access
Memory, can also include nonvolatile memory, a for example, at least disk memory, flush memory device or other are volatile
Property solid-state memory.
The present embodiment additionally provides a kind of computer storage media, for storing the computer journey used in above-mentioned camera
Sequence.
In several embodiments provided herein, it should be understood that disclosed device and method can also pass through
Other modes are realized.The apparatus embodiments described above are merely exemplary, for example, the flow chart in attached drawing and structure
Figure show the device of multiple embodiments according to the present invention, method and computer program product system frame in the cards
Structure, function and operation.In this regard, each box in flowchart or block diagram can represent a module, section or code
A part, the part of the module, section or code includes one or more for implementing the specified logical function
Executable instruction.It should also be noted that in the realization method as replacement, the function of being marked in box can also be to be different from
The sequence marked in attached drawing occurs.For example, two continuous boxes can essentially be basically executed in parallel, they are sometimes
It can execute in the opposite order, this is depended on the functions involved.It is also noted that in structure chart and/or flow chart
The combination of each box and the box in structure chart and/or flow chart can use the special of function or action as defined in executing
Hardware based system realize, or can realize using a combination of dedicated hardware and computer instructions.
In addition, each function module or unit in each embodiment of the present invention can integrate and to form an independence
Part, can also be modules individualism, can also two or more modules be integrated to form an independent part.
It, can be with if the function is realized and when sold or used as an independent product in the form of software function module
It is stored in a computer read/write memory medium.Based on this understanding, technical scheme of the present invention is substantially in other words
The part of the part that contributes to existing technology or the technical solution can be expressed in the form of software products, the meter
Calculation machine software product is stored in a storage medium, including some instructions are used so that a computer equipment (can be intelligence
Can mobile phone, personal computer, server or network equipment etc.) execute each embodiment the method for the present invention whole or
Part steps.And storage medium above-mentioned includes：USB flash disk, mobile hard disk, read-only memory (ROM, Read-Only Memory),
Random access memory (RAM, Random Access Memory), magnetic disc or CD etc. are various can to store program code
Medium.
The above description is merely a specific embodiment, but scope of protection of the present invention is not limited thereto, any
Those familiar with the art in the technical scope disclosed by the present invention, can easily think of the change or the replacement, and should all contain
Lid is within protection scope of the present invention.Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (10)
1. a kind of camera self-adapting regulation method, which is characterized in that including：
The extensive chemical that the adjustment action input of the kernel state information of multiple determinations and corresponding initial policy is pre-established
Model is practised, intensified learning parameter is generated；
When adjust automatically, using current kernel state information, the intensified learning model and the intensified learning parameter, obtain
Optimal correction is taken to act.
2. camera self-adapting regulation method according to claim 1, which is characterized in that the intensified learning model is
On-policy linear sarsa intensified learning models.
3. camera self-adapting regulation method according to claim 1, which is characterized in that the intensified learning model carries out
The action value function of fitting is：
Above formula S is the kernel state information, and A acts for the adjustment, and w is the intensified learning parameter, and x is learning sample,
I.e. To strengthen the gradient of parameter w.
4. camera self-adapting regulation method according to claim 3, which is characterized in that the institute of the intensified learning model
The object function for stating intensified learning parameter w is：
Above formula qπ(S, A) is the true action value function based on the initial policy, using based on sarsa (λ) algorithmIn advance
Estimate：
Above formula Rt+nThe gain brought is acted for adjustment described in the t+n moment, λ is the preset hyper parameter of sarsa (λ) algorithm, and γ is
The preset hyper parameter of intensified learning, Q (St+n) beπ is the initial policy.
5. camera self-adapting regulation method according to claim 1, which is characterized in that further include：
According to intensified learning parameter described in optimal correction action and the current kernel state information updating and optimizing.
6. camera self-adapting regulation method according to claim 5, which is characterized in that the update optimizes the reinforcing
The formula of learning parameter is：
Above formulaTo strengthen the gradient of parameter w.
7. a kind of camera self-adapting adjusting apparatus, which is characterized in that including：
Parameter generation module is used for the adjustment action input of the kernel state information of multiple determinations and corresponding initial policy
The intensified learning model pre-established generates intensified learning parameter；
Automatic regulating module utilizes current kernel state information, the intensified learning model and institute when being used for adjust automatically
Intensified learning parameter is stated, optimal correction action is obtained.
8. camera self-adapting adjusting apparatus according to claim 7, which is characterized in that further include：
Parameter update module, for according to optimal correction action and the current kernel state information updating and optimizing institute
State intensified learning parameter.
9. a kind of camera, which is characterized in that including memory and processor, the memory is for storing computer journey
Sequence, the processor runs the computer program so that the camera is executed according to described in any one of claim 1 to 6
Camera self-adapting regulation method.
10. a kind of computer storage media, which is characterized in that it is stored with used in the camera described in claim 9
Computer program.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810614204.2A CN108803328B (en) | 2018-06-14 | 2018-06-14 | Camera self-adaptive adjusting method and device and camera |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810614204.2A CN108803328B (en) | 2018-06-14 | 2018-06-14 | Camera self-adaptive adjusting method and device and camera |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108803328A true CN108803328A (en) | 2018-11-13 |
| CN108803328B CN108803328B (en) | 2021-11-09 |
Family
ID=64086936
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810614204.2A Active CN108803328B (en) | 2018-06-14 | 2018-06-14 | Camera self-adaptive adjusting method and device and camera |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108803328B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112084925A (en) * | 2020-09-03 | 2020-12-15 | 厦门利德集团有限公司 | Intelligent electric power safety monitoring method and system |
| CN112734759A (en) * | 2021-03-30 | 2021-04-30 | 常州微亿智造科技有限公司 | Method and device for determining trigger point of flying shooting |
| CN113568305A (en) * | 2021-06-10 | 2021-10-29 | 贵州恰到科技有限公司 | Control method of deep reinforcement learning model robot |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2010045272A1 (en) * | 2008-10-14 | 2010-04-22 | Honda Motor Co., Ltd. | Smoothed sarsa: reinforcement learning for robot delivery tasks |
| WO2016061724A1 (en) * | 2014-10-20 | 2016-04-28 | 中国科学院自动化研究所 | All-weather video monitoring method based on deep learning |
| CN105549384A (en) * | 2015-09-01 | 2016-05-04 | 中国矿业大学 | Inverted pendulum control method based on neural network and reinforced learning |
| CN106483852A (en) * | 2016-12-30 | 2017-03-08 | 北京天恒长鹰科技股份有限公司 | A kind of stratospheric airship control method based on Q Learning algorithm and neutral net |
| CN106766006A (en) * | 2017-02-21 | 2017-05-31 | 华南理工大学 | Air-conditioning system adaptive temperature compensation device and method based on machine vision |
| CN107272785A (en) * | 2017-07-19 | 2017-10-20 | 北京上格云技术有限公司 | A kind of electromechanical equipment and its control method, computer-readable medium |
-
2018
- 2018-06-14 CN CN201810614204.2A patent/CN108803328B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2010045272A1 (en) * | 2008-10-14 | 2010-04-22 | Honda Motor Co., Ltd. | Smoothed sarsa: reinforcement learning for robot delivery tasks |
| WO2016061724A1 (en) * | 2014-10-20 | 2016-04-28 | 中国科学院自动化研究所 | All-weather video monitoring method based on deep learning |
| CN105549384A (en) * | 2015-09-01 | 2016-05-04 | 中国矿业大学 | Inverted pendulum control method based on neural network and reinforced learning |
| CN106483852A (en) * | 2016-12-30 | 2017-03-08 | 北京天恒长鹰科技股份有限公司 | A kind of stratospheric airship control method based on Q Learning algorithm and neutral net |
| CN106766006A (en) * | 2017-02-21 | 2017-05-31 | 华南理工大学 | Air-conditioning system adaptive temperature compensation device and method based on machine vision |
| CN107272785A (en) * | 2017-07-19 | 2017-10-20 | 北京上格云技术有限公司 | A kind of electromechanical equipment and its control method, computer-readable medium |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112084925A (en) * | 2020-09-03 | 2020-12-15 | 厦门利德集团有限公司 | Intelligent electric power safety monitoring method and system |
| CN112734759A (en) * | 2021-03-30 | 2021-04-30 | 常州微亿智造科技有限公司 | Method and device for determining trigger point of flying shooting |
| CN113568305A (en) * | 2021-06-10 | 2021-10-29 | 贵州恰到科技有限公司 | Control method of deep reinforcement learning model robot |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108803328B (en) | 2021-11-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113534936B (en) | Server fan rotating speed control method, device, equipment and medium | |
| CN108803328A (en) | Camera self-adapting regulation method, device and camera | |
| US8582807B2 (en) | Systems and methods for determining personal characteristics | |
| Nakanishi et al. | Composite adaptive control with locally weighted statistical learning | |
| US9395708B2 (en) | Systems and methods for adaptive sampling rate adjustment | |
| EP3579187A1 (en) | Facial tracking method, apparatus, storage medium and electronic device | |
| Ling | Computational intelligence and its applications: evolutionary computation, fuzzy logic, neural network and support vector machine techniques | |
| CN104243835A (en) | Automatic diaphragm control method and system | |
| US11270203B2 (en) | Apparatus and method for training neural network by performing normalization using a plurality of normalization techniques | |
| CN109818971B (en) | Network data anomaly detection method and system based on high-order association mining | |
| CN111860789A (en) | Model training method, terminal and storage medium | |
| CN108691795B (en) | The fan control of electronic equipment and modification method | |
| CN112232426A (en) | Training method, device and equipment of target detection model and readable storage medium | |
| CN109139524A (en) | A kind of method, apparatus and storage medium of fan speed regulation | |
| CN113801967A (en) | Method and device for regulating and controlling temperature of incubator and incubator | |
| Cui et al. | Neural-network-based distributed adaptive synchronization for nonlinear multi-agent systems in pure-feedback form | |
| CN113960923A (en) | Model-free self-adaptive sliding mode control method based on discrete extended state observer | |
| CN116560475A (en) | Server fan control method and computer equipment | |
| GB2627119A (en) | Recommendation generation using machine learning data validation | |
| CN116050506A (en) | Back door attack defense method and device, electronic equipment and storage medium | |
| CN117994635A (en) | Federal element learning image recognition method and system with enhanced noise robustness | |
| CN112116024A (en) | Method and device for classifying models by user, electronic equipment and storage medium | |
| CN111708748B (en) | Management system and method of weblog analysis algorithm | |
| CN102222340A (en) | Method and system for detecting prospect | |
| CN113638900A (en) | Method, device and equipment for regulating and controlling storage array fan and readable medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |