CN110429652A - A kind of intelligent power generation control method for expanding the adaptive Dynamic Programming of deep width - Google Patents
A kind of intelligent power generation control method for expanding the adaptive Dynamic Programming of deep width Download PDFInfo
- Publication number
- CN110429652A CN110429652A CN201910801884.3A CN201910801884A CN110429652A CN 110429652 A CN110429652 A CN 110429652A CN 201910801884 A CN201910801884 A CN 201910801884A CN 110429652 A CN110429652 A CN 110429652A
- Authority
- CN
- China
- Prior art keywords
- width
- tri
- network
- dynamic programming
- deep
- 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
Abstract
The present invention provides a kind of intelligent power generation control method that can expand the adaptive Dynamic Programming of deep width, this method can expand the adaptive dynamic programming algorithm of deep width in conjunction with learning rate adaptive algorithm, be suitable for the tri-state energy real power control of " i.e. source, that is, lotus " under following comprehensive energy interacted system (FIEIS) unified time scale.This method solve electric system under the tri-state energy and the high permeability of intermittent energy source, and conventional electric power generation controls the real-time flexible control problem for being difficult to adapt to active frequency, has higher robustness and stability, and iterative convergence speed and accuracy rate get a promotion.According to the depth and width of the tri-state energy source type quantity adaptively changing deep neural network of access network on this method frame, to enhance the real-time of real power control, Stability and veracity.Deep width model prediction network, evaluation network and the execution network of mentioned method, effectively instead of traditional Multiple Time Scales combined type real power control algorithm.
Description
Technical field
The present invention relates to the intelligent power generation control methods that one kind can expand the adaptive Dynamic Programming of deep width, belong to power train
Intelligent power generation of uniting is dispatched and control field, is applied to " i.e. source, that is, lotus " under the following comprehensive energy interacted system unified time scale
Tri-state energy active power and frequency control.
Background technique
As the intermittent new energy permeability of wind-power electricity generation and photovoltaic power generation etc. rises, the economy of electric network active frequency is excellent
Change configuration and stable regulation faces the challenge.To ensure that power grid security economical operation solves the problems, such as the active frequency difference of output, in addition to
Active power output is adjusted with conventional Power Generation Modes such as water power and thermoelectricitys, electric car, power battery and flywheel energy storage can also be utilized
Come that partial equilibrium system loading is active and frequency fluctuation in the form of charge and discharge Deng the tri-state energy access power grid with energy-storage function.
When introducing electric car and power battery etc., which participate in the active frequency of system, to be adjusted, sending out active has at random not
Certainty and " i.e. source, that is, lotus " characteristic, this increases difficulty to the regulation of power grid.By taking electric car participates in the active adjusting of system as an example,
Its active stochastic uncertainty is embodied in the use of electric car by the more of user's habit, weather conditions and festivals or holidays trip etc.
Weight factor influences；" i.e. source, that is, the lotus " of electric car then refers to that electric car can be used as power generation end state, electricity consumption end state and shut down
State, i.e. the tri-state energy, and its state is real-time change.In addition to the tri-state energy affects electric network active, wind-powered electricity generation, tide
Nighttide, which can generate electricity, will also will increase power grid electric energy real power control burden with the extensive access of the intermittent new energy such as photovoltaic power generation.
To solve the real-time adjusting control that distributed new accesses electric network active frequency, risen especially for industry is adapted to
For grade to the requirements at the higher level of power quality, the following smart grid, which is built, will can accommodate more tri-state energy.Smart grid
It should be the sturdy power grid that economy, reliability and flexibility are had both.In the algorithm field for realizing electric power system optimization control, expert
Scholars propose many solutions.The power-frequency distribution method of deep learning solution is such as utilized, for another example adaptive width
The intelligent power generation method of Dynamic Programming, these methods solve the active adjusting of system to a certain extent.But in face of increasingly complexity
Power system environment is more with tradition especially under the high permeability complex environment of the following tri-state energy and distributed new
The method of time scale carries out the problem of control can bring lag.Therefore, there is still a need for seeking one kind has stronger learning ability,
The better algorithm of robustness has come the tri-state Integrated Energy for solving following comprehensive energy interacted system (FIEIS) the unified time scale
Function control problem.
Summary of the invention
The present invention proposes that a kind of intelligent power generation control method that can expand the adaptive Dynamic Programming of deep width, this method are endeavoured
Active and frequency high-quality control is difficult to realize in a manner of conventional ASSEMBLE Generation Control after solving FIEIS and generating electricity by way of merging two or more grid systems to ask
Topic.Tradition based on the combined power generation schduling control algorithm of Multiple Time Scales scheduling controlling after tri-state energy scale is grid-connected, function
The fluctuation of rate randomness causes frequency departure that cannot timely regulate and control.The Automatic Generation Control of unified time scale is with 4s for one
Period avoids Unit Combination caused by the scheduling of script Multiple Time Scales and economic load dispatching distribution delay.
It is proposed by the present invention based on the control method that can expand the adaptive Dynamic Programming of deep width, solving the active frequency that generates electricity
Rate is adjusted on control frame used, the neural network of different previous adaptive Dynamic Programming (ADP) algorithms.But with depth mind
Through the prolongable algorithm frame of depth and width based on network, network is stacked with several limited Boltzmann machines (RBM)
It forms.Include visible layer and hidden layer in limited Boltzmann machine, in identical layer between unit it is connectionless.Limited Boltzmann
Function flow function is as follows:
R in formulaiFor the biasing of visible layer unit i, cjFor the biasing for hiding layer unit j, LijIt is connect for unit i with unit j
Weight, ρ are the parameter sets of all connection weights and biasing, viFor visible layer unit i, hjTo hide layer unit j, with vi×hj
Indicate viWith hjCorrelation degree.The joint probability distribution of state variable (v, h) indicates are as follows:
F (v, h | ρ)=exp-E (v, h | ρ) }/z (2)
In formula
It needs to take its logarithm using to likelihood function for convenience of calculating in limited Boltzmann machine training process:
Local derviation is asked to calculate gradient to above formula:
Then updating unit biases ri、cjAnd connection weight LijEtc. parameters:
Local derviation indicates in above formula are as follows:
After the parameter calculating for completing each layer, the activation condition of visible layer unit is then sought, to activate probability to indicate:
σ is activation primitive in formula, and to avoid trained gradient disappearance problem, ReLU activation can be used in deep neural network
Linear classification and problem concerning study, expression formula are Function Solution by no meansThe activation for hiding layer unit is general
Rate then indicates are as follows:
The probability of movement is converted are as follows:
K in formulaaExpression acts probability coefficent, the state conversion of probability are as follows:
T in formulasIndicate probability transformation ratio.
Based on the intelligent power generation control method for the adaptive Dynamic Programming that can expand deep width, original state does not have net
Network predictive ability needs to carry out off-line training in advance for the accurate system operation characteristic that obtains.Off-line training is coveted using unsupervised
Greedy layer-by-layer training method is trained the passing data sample for having calibration and obtains the initial value of network weight.To guarantee algorithm
Learn correctness, these samples should be the inputoutput data that can reflect the true operating status of passing system comprehensively.Next is
Sample over-fitting is prevented, the technology that such as Dropout can be used improves grid generalization ability.On-line control system power generation
When, using the training method training real time data for having supervision and off-line training result is finely tuned, to constantly update learning outcome.
The expansion of adaptive its width of Dynamic Programming electricity-generating method and depth that can expand deep width depends on access electric power
The tri-state energy of grid and the quantity of distributed energy.As previously mentioned, there is power generation, electricity consumption in the tri-state energy of access system
With the state of shutting down.According to the different conditions of the tri-state energy and access system distributed energy start and stop state, on-line tuning depth mind
Depth and width through network, to improve the accuracy for obtaining system features and forecasting system operating parameter.
Detailed description of the invention
Fig. 1 is that the method for the present invention expands depth and width neural network schematic diagram.
Fig. 2 is the intermittent new energy and tri-state energy power generation curve synoptic diagram of the method for the present invention.
Fig. 3 is that the method for the present invention is applied to unified time scale Generation Control mode block schematic illustration.
Specific embodiment
One kind proposed by the present invention can expand the intelligent power generation control method of the adaptive Dynamic Programming of deep width, in conjunction with attached drawing
Detailed description are as follows:
Fig. 1 is that the method for the present invention expands depth and width neural network schematic diagram.Deep width neural network includes deep
Width model predicts that network, deep width evaluation network and deep width execute network.These network structures are similar, are only defeated in network
Enter and export different from content.It includes input layer, output layer and 2 hidden layers that deep width neural network is expanded in this figure.
With m in figurei, m0Indicate the neural unit number of input layer and output layer；With m1、m2、…、mjIndicate each hidden layer unit number.When
The access system tri-state energy or distributed energy quantity change, and depth model predicts that network is to adapt to access behavior aggregate and shape
State number increases input layer unit number and expands width, and network is by original { mi,m1,m2,m0Widen as { mi+Δmi,2(m1+Δ
mi),m2,m0}；Depth, which executes network, can then expand output layer unit number, and network is by original { mi,m1,m2,m0Widen as { mi,
m1,2(m2+Δmo),mo+Δmo, increase output action quantity feedback in access topology node with control the active of electric system and
Frequency departure.For the increase of adaptive learning complexity, deep width neural network adaptively expands network depth, and former network can be changed to
{mi,m1,m2,m3,m4,m5,mo, or change simultaneously the depth and width of network.
Fig. 2 is the intermittent new energy and tri-state energy power generation curve synoptic diagram of the method for the present invention.It is in figure it can be seen that all
Such as wind-power electricity generation, photovoltaic power generation and the tidal power energy have the characteristics that intermittence, and different power generation type quality are also variant,
And it is contemplated that the following power grid intermittence new energy permeability will be promoted further.As shown in the figure, the tri-state energy has electricity
Source, load and shut down three states.It, can be flat using it after accessing power grid such as electric car, energy storage battery and pumped-storage power generation
Balance system is active and adjusts frequency, plays the role of peak load shifting.Either intermittent new energy or the tri-state energy, connect on a large scale
Entering rear electric power system dispatching control will become complicated, and especially it, will with conventional electric power generation control mode with intermittent part
It is difficult to adjust in real time.The demand for development of the following smart grid can accommodate frequency and active after the access of diversified forms power generation energy resource
Stable regulation.
Fig. 3 is that the method for the present invention is applied to unified time scale Generation Control mode block schematic illustration.Consider not in frame
After coming the power grid tri-state energy and the raising of distributed new permeability, problem is uniformly regulated and controled to the power generation in part or region.It is deep wide
Neural network successively training offline in advance is spent, depth recognizes each topological node feature of the power grid of the energy containing tri-state.For isolated power grid
When network topology structure and power supply form the special screne that varies widely, need to reinforce the pre-training of intelligent body, with reality
Now power grid control is more intelligent, and system is more stable.When this method is used for real system on-line training, from grid topology
Conduct can expand deep width neural network after the transformed processing of characteristic of Node extraction hair electricity condition, power and frequency etc.
Input state parameter matrix S=[S1,S2,S3,…,Sm].WhereinIndicate i-th unit (including
The tri-state energy) k of input can characterize the state parameter of unit generation characteristic.Net is executed from deep width through intelligent body internal calculation
Network output action matrix A=[A1,A2,A3,…,Am], whereinIndicate i-th unit (including tri-state
The energy) generate k instruct.A acts matrix as control instruction, adjusts power source combination generated output and frequency deviation f simultaneously
Control tri-state energy state；A matrix is supplied to intelligent body on-line training as sample simultaneously, exports control result to subsequent time
Fine tuning.Power supply and the increase of tri-state energy source type quantity when access network, the automatic width and depth for expanding deep width neural network
Degree, to adapt to more complicated network.
Claims (5)
1. the intelligent power generation control method that one kind can expand the adaptive Dynamic Programming of deep width, which is characterized in that can be by automatic
Deep neural network depth and width are expanded to adapt to network topology structure variation；It is embodied in when the access power grid tri-state energy
Number of types changes, and algorithm adjust automatically learns depth, that is, deep neural network number of plies to adapt to system structure complexity；Automatically
The width of regularized learning algorithm, that is, deep neural network neuron number with adapt to access network topology variation；It can adapt to the tri-state energy
To the control of active randomness fluctuation after access system, and will be with unified dispatch command and the smallest delayed-action in intelligence
Electricity generation system；Different from the adaptive dynamic programming algorithm of conventional depth, the adaptive dynamic programming algorithm of deep width can be expanded by mentioning
By training system history run parameter, adjust automatically learning rate realize to the different classes of energy of input system it is accurate response with
It quickly calculates, controls especially for the randomness load access electric power networks of the tri-state energy of " i.e. source, that is, lotus " with stronger
Robustness and stability.
2. one kind as described in claim 1 can expand the intelligent power generation control method of the adaptive Dynamic Programming of deep width, special
Sign is that this method frame breaches previous adaptive Dynamic Programming Generation Control mode neural network based, in deep width
It is subject to that deep width model prediction network can be expanded on the basis of adaptive dynamic programming algorithm；Strengthen neural network substitution with deep width
Deep neural network, so that system has stronger cognitive ability on generated output distribution Schistosomiasis control；By being controlled to power generation
Policy model processed improves, and this method is capable of handling increasingly complex network system, forecasting system subsequent time active state
Accuracy be improved.
3. one kind as described in claim 1 can expand the intelligent power generation control method of the adaptive Dynamic Programming of deep width, special
Sign is, can online adaptive expand study depth and width；Different from the adaptive dynamic programming algorithm of conventional depth, mentioning can
Online expansion network architecture can be changed with the data type of access system by expanding the adaptive dynamic programming algorithm of deep width；
By taking electric system as an example, when the grid structure of system changes, when the disconnection or access of power supply or the tri-state energy, percentage regulation
Learning network depth and width realize active automatic control by deep width learning system operation history data and network characterization
System.
4. one kind as described in claim 1 can expand the intelligent power generation control method of the adaptive Dynamic Programming of deep width, special
Sign is, can expand adaptive regularized learning algorithm rate and online updating learning strategy according to access system tri-state energy property；Such as electricity
Electrical automobile, power battery and the tri-state of the water-storage energy have the characteristics that active random fluctuation characteristic and " i.e. source i.e. lotus ", and each
Power-supply fluctuation characteristic is different；By offline pre-training, the network unit learning rate initial value that the tri-state energy participates in system is obtained, it can be with
Online adaptive adjustment algorithm learning rate is to obtain more excellent predicted state and faster rate of convergence and gradient is avoided to disappear；Equally
The adjustment of this feature learning rate be also applied for the Generation Control of micro-capacitance sensor and island-grid.
5. one kind as described in claim 1 can expand the intelligent power generation control method of the adaptive Dynamic Programming of deep width, special
Sign is, by power generation dispatching with unified time scale control instruction；Conventional electric power generation control generally with " Unit Combination+economic load dispatching+
The combined type control mode of Automatic Generation Control+power of the assembling unit distribution ", the Generation Control mode of this Multiple Time Scales will be not
Sluggishness is responded after carrying out extensive tri-state energy access；With the system real power control of the tri-state energy access when scheduling of unified time scale
Response will be more flexible and timely and unified scheduling without in addition increase machine unit scheduling instruction, thus reduce control again
Miscellaneous degree, increases system reliability.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910801884.3A CN110429652B (en) | 2019-08-28 | 2019-08-28 | Intelligent power generation control method capable of expanding deep width self-adaptive dynamic planning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910801884.3A CN110429652B (en) | 2019-08-28 | 2019-08-28 | Intelligent power generation control method capable of expanding deep width self-adaptive dynamic planning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110429652A true CN110429652A (en) | 2019-11-08 |
| CN110429652B CN110429652B (en) | 2021-03-26 |
Family
ID=68417855
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910801884.3A Active CN110429652B (en) | 2019-08-28 | 2019-08-28 | Intelligent power generation control method capable of expanding deep width self-adaptive dynamic planning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110429652B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110889780A (en) * | 2019-12-04 | 2020-03-17 | 浙江大学 | Wind turbine generator system performance-guaranteeing maximum wind energy capturing method based on BLS and double-layer SMC |
| CN111555297A (en) * | 2020-05-21 | 2020-08-18 | 广西大学 | Unified time scale voltage control method with tri-state energy unit |
| CN111769547A (en) * | 2020-06-12 | 2020-10-13 | 广西大学 | Real-time regulation and control method for three-layer linkage mechanism interactive comprehensive energy system |
| CN111799820A (en) * | 2020-05-27 | 2020-10-20 | 广西大学 | Double-layer intelligent hybrid zero-star cloud energy storage countermeasure regulation and control method for power system |
| CN112564118A (en) * | 2020-11-23 | 2021-03-26 | 广西大学 | Distributed real-time voltage control method capable of expanding quantum deep width learning |
| CN115690897A (en) * | 2022-08-31 | 2023-02-03 | 北京夕阳无忧科技有限公司 | Accidental behavior processing method, device, equipment and storage medium for preventing privacy leakage |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105790314A (en) * | 2016-03-08 | 2016-07-20 | 南京邮电大学 | Self-adaptive dynamic planning based coordinative control method of distributed power generator |
| CN105846461A (en) * | 2016-04-28 | 2016-08-10 | 中国电力科学研究院 | Self-adaptive dynamic planning control method and system for large-scale energy storage power station |
| CN109149648A (en) * | 2018-10-11 | 2019-01-04 | 广西大学 | A kind of adaptive width Dynamic Programming intelligent power generation control method |
| CN109193641A (en) * | 2018-10-11 | 2019-01-11 | 广西大学 | A kind of tri-state energy control method based on automatic expansion deep learning |
-
2019
- 2019-08-28 CN CN201910801884.3A patent/CN110429652B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105790314A (en) * | 2016-03-08 | 2016-07-20 | 南京邮电大学 | Self-adaptive dynamic planning based coordinative control method of distributed power generator |
| CN105846461A (en) * | 2016-04-28 | 2016-08-10 | 中国电力科学研究院 | Self-adaptive dynamic planning control method and system for large-scale energy storage power station |
| CN109149648A (en) * | 2018-10-11 | 2019-01-04 | 广西大学 | A kind of adaptive width Dynamic Programming intelligent power generation control method |
| CN109193641A (en) * | 2018-10-11 | 2019-01-11 | 广西大学 | A kind of tri-state energy control method based on automatic expansion deep learning |
Non-Patent Citations (4)
| Title |
|---|
| LINFEI YIN等: "Adaptive Distributed Consensus Protocol for Automatic Generation Control of Large-Scale Interconnected Power Systems", 《IEEE ACCESS》 * |
| 殷林飞: "基于深度强化学习的电力系统智能发电控制", 《中国博士学位论文全文数据库 工程科技》 * |
| 殷林飞等: "基于深度自适应动态规划的孤岛主动配电网发电控制与优化一体化算法", 《控制理论与应用》 * |
| 王曙: "深度学习算法研究及其在图像分类上的应用", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110889780A (en) * | 2019-12-04 | 2020-03-17 | 浙江大学 | Wind turbine generator system performance-guaranteeing maximum wind energy capturing method based on BLS and double-layer SMC |
| CN110889780B (en) * | 2019-12-04 | 2023-03-31 | 浙江大学 | Wind turbine generator system performance-guaranteeing maximum wind energy capturing method based on BLS and double-layer SMC |
| CN111555297A (en) * | 2020-05-21 | 2020-08-18 | 广西大学 | Unified time scale voltage control method with tri-state energy unit |
| CN111555297B (en) * | 2020-05-21 | 2022-04-29 | 广西大学 | Unified time scale voltage control method with tri-state energy unit |
| CN111799820A (en) * | 2020-05-27 | 2020-10-20 | 广西大学 | Double-layer intelligent hybrid zero-star cloud energy storage countermeasure regulation and control method for power system |
| CN111799820B (en) * | 2020-05-27 | 2022-07-05 | 广西大学 | Double-layer intelligent hybrid zero-star cloud energy storage countermeasure regulation and control method for power system |
| CN111769547A (en) * | 2020-06-12 | 2020-10-13 | 广西大学 | Real-time regulation and control method for three-layer linkage mechanism interactive comprehensive energy system |
| CN111769547B (en) * | 2020-06-12 | 2023-07-28 | 广西大学 | Real-time regulation and control method for interactive comprehensive energy system of three-layer linkage mechanism |
| CN112564118A (en) * | 2020-11-23 | 2021-03-26 | 广西大学 | Distributed real-time voltage control method capable of expanding quantum deep width learning |
| CN115690897A (en) * | 2022-08-31 | 2023-02-03 | 北京夕阳无忧科技有限公司 | Accidental behavior processing method, device, equipment and storage medium for preventing privacy leakage |
| CN115690897B (en) * | 2022-08-31 | 2023-10-20 | 北京夕阳无忧科技有限公司 | Unexpected behavior processing method, device and equipment for preventing privacy leakage and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110429652B (en) | 2021-03-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110429652A (en) | A kind of intelligent power generation control method for expanding the adaptive Dynamic Programming of deep width | |
| Li et al. | Coordinated load frequency control of multi-area integrated energy system using multi-agent deep reinforcement learning | |
| Wen et al. | Real-time identification of power fluctuations based on LSTM recurrent neural network: A case study on Singapore power system | |
| CN111884213B (en) | Power distribution network voltage adjusting method based on deep reinforcement learning algorithm | |
| CN112186743B (en) | Dynamic power system economic dispatching method based on deep reinforcement learning | |
| CN112614009B (en) | Power grid energy management method and system based on deep expectation Q-learning | |
| CN111934335A (en) | Cluster electric vehicle charging behavior optimization method based on deep reinforcement learning | |
| CN103729695A (en) | Short-term power load forecasting method based on particle swarm and BP neural network | |
| Chitsazan et al. | Wind speed forecasting using an echo state network with nonlinear output functions | |
| CN106532691A (en) | Adaptive dynamic programming-based frequency compound control method of single-region power system | |
| Georgiou et al. | Implementing artificial neural networks in energy building applications—A review | |
| CN109193641A (en) | A kind of tri-state energy control method based on automatic expansion deep learning | |
| US10686327B2 (en) | Energy storage controller | |
| Yin et al. | Mode-decomposition memory reinforcement network strategy for smart generation control in multi-area power systems containing renewable energy | |
| CN115795992A (en) | Park energy Internet online scheduling method based on virtual deduction of operation situation | |
| CN117039981A (en) | Large-scale power grid optimal scheduling method, device and storage medium for new energy | |
| CN111799808A (en) | Power grid reactive voltage distributed control method and system | |
| Yin et al. | Expandable deep width learning for voltage control of three-state energy model based smart grids containing flexible energy sources | |
| Yin et al. | Expandable depth and width adaptive dynamic programming for economic smart generation control of smart grids | |
| Li et al. | Data‐driven cooperative load frequency control method for microgrids using effective exploration‐distributed multi‐agent deep reinforcement learning | |
| Tao et al. | On comparing six optimization algorithms for network-based wind speed forecasting | |
| CN111799820B (en) | Double-layer intelligent hybrid zero-star cloud energy storage countermeasure regulation and control method for power system | |
| Vohra et al. | End-to-end learning with multiple modalities for system-optimised renewables nowcasting | |
| CN116865343A (en) | Model-free self-adaptive control method, device and medium for distributed photovoltaic power distribution network | |
| Prema et al. | LSTM based Deep Learning model for accurate wind speed prediction |
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 |