CN105005866A - Robustness assessing method of power system reactive optimization control policy - Google Patents
Robustness assessing method of power system reactive optimization control policy Download PDFInfo
- Publication number
- CN105005866A CN105005866A CN201510461517.5A CN201510461517A CN105005866A CN 105005866 A CN105005866 A CN 105005866A CN 201510461517 A CN201510461517 A CN 201510461517A CN 105005866 A CN105005866 A CN 105005866A
- Authority
- CN
- China
- Prior art keywords
- robustness
- performance index
- reactive power
- control strategy
- variable
- 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.)
- Pending
Links
Landscapes
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention discloses a robustness assessing method of a power system reactive optimization control policy. A random sampling method is used to acquire the best performance index and the worst performance index in a global space and the best performance index and the worst performance index of a robust space composed of the controllable accuracy range of various control variables and the change range of various state variables in the vicinity of a specific reactive optimization control policy. On the basis, a robustness index and an assessing policy for the specific reactive optimization control policy are defined. On the basis of the optimization result, a decision variable which can withstand a certain degree of parameter fluctuation is emphasized and searched. The policy with excellent robustness is selected and applied to a project. The shortcoming that small perturbations of control variable and state variable loads and other parameters are often ignored by a reactive optimization scheme is effectively compensated.
Description
Technical field
The present invention relates to reactive power optimization of power system field, particularly a kind of robustness appraisal procedure of reactive power optimization of power system control strategy.
Background technology
Traditional idle work optimization research great majority, based on deterministic models, namely suppose that load level, control variable numerical value etc. remain unchanged, are putting the idle regulation and control scheme that Optimal Decision-making improves some operating index before this.Operation of Electric Systems often faces a little uncertain problem: the controlled accuracy limitations of continuous type control variable, completely by control accurately, can not there is stochastic error; State variable load moment changes, even if section also exists predicated error etc. sometime.If the uncertain disturbances of Reactive power control strategy to control variable and state variable is very responsive, then its implementation result may not be good, even causes the infeasible method of operation or security of system stability state to worsen.
Compared to determinacy idle work optimization lot of documents and enrich achievement, consider that the idle work optimization research of uncertain factor is still in stage of initial development.
The prioritization scheme that the idle work optimization bilayer model that document " information uncertainty is on the impact of reactive power optimization " proposes draws has more redundance than scheme when not taking into account load change.
Document " is taken into account the idle work optimization model and algorithl of negative rules " and is proposed a kind of idle work optimization model and algorithl taking into account the influence factors such as negative rules, to overcome the deficiency that idle work optimization model under single load level can not portray its load stochastic characteristic comprehensively.
Document " Stochastic optimal reactive power dispatch:Formulation and solution method " proposes the random idle work optimization Chance-Constrained Programming Model considering the uncertain factor such as load and branch trouble, and the optimum results of gained improves its feasibility under Uncertain environments on the basis of sacrificing to a certain degree objective function.
Document " takes into account the rural power grids idle work optimization method of negative rules ", and consideration affects the uncertain factor of load statistics, and propose to set up three ladder load tdistribution curve models, optimum results shows that the method falls damage successful when load change.
Document " Non-dominated sorting genetic algorithm-II for robust multi-objective optimalreactive power dispatch " adopts the Robust Optimization Model of monte carlo integration form, but its load sample methods of sampling proposed needs prior given system power growing direction, and thus the robust performance of optimum results exists a definite limitation.
Summary of the invention
The object of the present invention is to provide a kind of robustness appraisal procedure of reactive power optimization of power system control strategy, on the basis of its optimum results, emphasize to search for the decision variable can resisting to a certain degree parameter fluctuation, select the excellent application of policies of robustness in engineering, effectively compensate for idle work optimization scheme and often ignore this drawback of parameter microvariations such as control variable and state variable load.
To achieve these goals, the present invention adopts following technical scheme:
A robustness appraisal procedure for reactive power optimization of power system control strategy, comprises the following steps:
1) be global space by combining by the span of each control variable and the variation range of each state variable the definition space formed in electric system to be optimized; By near the Reactive power control strategy C of certain pending robustness assessment, combine by the controlled accuracy rating of each control variable and the variation range of each state variable the space formed, be defined as robust space;
2) for single performance index, employing random sampling optimization method obtains the optimal performance index f in global space
b, the poorest performance index are f
w, and Reactive power control strategy C robust space to be assessed in optimal performance index f
b,c, the poorest performance index are f
w,c, corresponding minimum sampling number N
min, computing formula is:
For global space, the value of p% is 0.1% ~ 0.5%; The value of q% is 0.5% ~ 1%;
For robust space, the value of p% is 1% ~ 2%; The value of q% is 0.5% ~ 1%;
3) in global space, optimal performance index with the difference Δ f of the poorest performance index is:
Δf＝f
b-f
w
For Reactive power control strategy C to be assessed, in its robust space, the difference Δ f of optimal performance index and the poorest performance index
cfor:
Δf
c＝f
b,c-f
w,c
The robustness index R of definition control strategy C
cfor:
4) if R
c> threshold value R
set, then think that the robustness of Reactive power control strategy C meets the demands.Threshold value R
setbe more than or equal to 0.8, rule of thumb can be set by appraiser.
The present invention further improves and is: for the situation of multi-performance index, when only meeting the demands for the robustness of each performance index, thinks that the robustness of this Reactive power control strategy C meets the demands.
The present invention further improves and is: the situation several Reactive power control strategy all being met to robustness requirement, the mode taking equal proportion to expand the control accuracy of each control variable and the variation range of state variable gradually expands robust space, and re-start robustness assessment, do not meet the Reactive power control strategy of robustness requirement after the robustness first not meeting the Reactive power control strategy of robustness requirement is weaker than.
The present invention further improves and is: the situation several idle work optimization strategy all not being met to robustness requirement, by the control accuracy of each control variable and the equal scaled down of the variation range of state variable to β %, if the robustness within the scope of this meets the demands, then the receptance of the robustness of this idle work optimization strategy is claimed to be β %.
Relative to prior art, the present invention has following beneficial effect:
1) the present invention effectively compensate for idle work optimization scheme and often ignores this drawback of parameter microvariations such as control variable and state variable load.
2) the present invention considers the realized precision of control variable and the uncertainty of state variable, establishes the index assessed the robustness of control strategy, has both been applicable to single performance index optimization problem and has also been applicable to multi-performance index optimization problem.
Accompanying drawing explanation
Fig. 1 is IEEE-30 node modular system figure.
Embodiment
The robustness appraisal procedure of a kind of reactive power optimization of power system control strategy of the present invention, comprises the following steps:
1) be global space by combining by the span of each control variable and the variation range of each state variable the definition space formed in electric system to be optimized; By near the Reactive power control strategy of certain pending robustness assessment, combine by the controlled accuracy rating of each control variable and the variation range of each state variable the space formed, be defined as robust space.
Control variable refers to the control object that may be used for carrying out reactive power optimization of power system, comprising: each grouping switching shnt capacitor, each ULTC, each generator voltage etc.
State variable mainly refers to the load of each node.
2) (establish these performance index to be the bigger the better) when single performance index, the optimal performance index be located in global space is f
b, the poorest performance index are f
w, both difference Δ f are:
Δf＝f
b-f
w
For a concrete Reactive power control strategy C, in its robust space, suppose that its optimal performance index is f
b,c, the poorest performance index are f
w,c, both difference Δ f
cfor:
Δf
c＝f
b,c-f
w,c
The robustness index R of definition Reactive power control strategy C
cfor:
3) a threshold value Rset is set according to actual needs, if Rc>Rset, then thinks that the robustness of Reactive power control strategy C meets the demands.
Threshold value can artificially set, and generally should be more than or equal to 0.8.
4) for the situation of multi-performance index, when only the robustness of each performance index being met the demands, just think that the robustness of Reactive power control strategy C meets the demands.
5) several Reactive power control strategy is all met to the situation of robustness requirement, the mode taking equal proportion to expand the control accuracy of each control variable and the variation range of state variable gradually expands robust space, and re-start robustness assessment, do not meet the Reactive power control strategy of robustness requirement after the robustness first not meeting the Reactive power control strategy of robustness requirement is weaker than.Several control strategy is not all met to the situation of robustness requirement, by the control accuracy of each control variable and the equal scaled down of the variation range of state variable to β %, if the robustness within the scope of this meets the demands, the receptance of the robustness of this Reactive power control strategy is then claimed to be β %, also can according to requirement of engineering, in conjunction with the controllability of each control variable and the uncertainty of each state variable, the variation range of the control accuracy of each control variable and state variable is carried out differentiation but not scaled down.
Optimal performance index f in global space
b, the poorest performance index are f
w, optimal performance index f in robust space
b,c, the poorest performance index are f
w,crandom sampling optimization method is all adopted to obtain, corresponding minimum sampling number N
min, computing formula is:
Because the individual amount in global space is more, p% generally desirable 0.1% ~ 0.5%, q% desirable 0.5% ~ 1%, namely confidence level is 99% ~ 99.5%, at it to each control variable and state variable may random sampling form and is no less than N in span
minindividual feasible sample, calculates the performance index of each sample respectively, therefrom draws optimal value and the worst-case value of performance index; Because the individual amount in robust space is less, p% generally desirable 1% ~ 2%, q% desirable 0.5% ~ 1%, namely confidence level is 99% ~ 99.5%, various discrete control variable is fixed as value corresponding in control strategy to be assessed, and random sampling formation in the span that its robust space is corresponding is no less than N to each stepless control variable and state variable
minindividual feasible sample, calculates the property indices of each sample respectively, therefrom draws optimal value and the worst-case value of performance index.
Below with reference to the IEEE30 node idle work optimization example shown in Fig. 1, the robustness appraisal procedure that the present invention proposes is described further.In IEEE30 node system, there are 41 branch roads, 6 generator nodes and 21 load buses.6 generator nodes are respectively 1,2,5,8,11,13 nodes (as balance node, all the other nodes are PV node to its interior joint 1), and in system, other node is PQ node, and this system basic parameter and service data are as shown in table 1-table 3:
Table 1 IEEE30 node system bus data and power flow solutions
Table 2 IEEE30 node system branch data (perunit value)
Adjustable bus data that table 3 is idle
The control object of this example comprises: grouping switching shnt capacitor (node 10 and 24, control variable name is defined as C1 and C2 respectively), ULTC (node 6-9, between 6-10,4-12,27-28, control variable name is defined as T1, T2, T3, T4 respectively), generator voltage (node 1,2,5,8,11,13, control variable name is defined as V1, V2, V3, V4, V5, V6 respectively), the information of each control variable is as shown in table 4.
Each control variable (p.u.) in table 4 example
On the basis of load data, superpose the uncertain parameters considering its fluctuation and predicated error thereof, 21 load buses are divided into municipal household electricity, tertiary industry electricity consumption and heavy industry trade power consumption 3 type, before load, 1-7 is the 1st class, load 8-14 is the 2nd class, load 15-21 is the 3rd class, and the stability bandwidth of 3 type load active power and reactive power is as shown in table 5 respectively.
The stability bandwidth of table 5 three type load active power and reactive power
If the predicated error of 3 type loads is followed successively by ± 1%, ± 0.5%, ± 1.5%, then consider the variation range number percent of each type load after its fluctuation and predicated error thereof as shown in table 6 respectively.
Table 6 considers the variation range number percent of each type load after fluctuation and predicated error thereof
With active power loss P
lossminimum is target, constraint condition comprises voltage deviation constraint, current limitation retrains and the range of control constraint of control variable, optimization method adopts sampling optimization method, select p% and q% to be 0.5%, namely confidence level is 99.5%, can be calculated, randomly draw 1058 samples and can draw satisfactory solution, carrying out 5 above-mentioned optimization respectively, showing that 5 satisfactory solutions are for carrying out robustness assessment, as shown in table 7.The solution that above-mentioned 5 suboptimization draw is added up, can show that objective function best values is f
b=0.06576, worst-case value is f
w=0.11895, Δ f=0.05319.
The control strategy (p.u.) of 5 pending robustness assessments that table 7 idle work optimization draws
Application sampling optimization method, samples in the robust space of each control strategy.When p% and q% all gets 1%, namely confidence level is 99%, calculates known, randomly draws 459 samples and can obtain required optimum and the poorest performance number, thus carry out robustness assessment.If threshold value R
setbe taken as 0.8, the robustness assessment result for 5 control strategies is as shown in table 8.
The robustness assessment result of table 85 control strategies
In conjunction with the threshold value R set
set, can be obtained by assessment result, second and third, the robustness of five groups of control strategies is undesirable, other two groups all meet the requirements.As single-object problem, can directly according to robustness index R
ccompare the robustness quality of the control strategy that respectively meets the demands, the robustness obtaining the 5th group of control strategy is optimum.For candidate's control strategy that limited performance index are close, can be passed through robustness assessment, the control strategy selecting robustness optimum is applied in engineering reality, and its ability adapting to various uncertain factor disturbance is stronger.By calculating the feasibility demonstrating robustness appraisal procedure.
If establish threshold value R
setwhen being 0.83, can be obtained by table 5, the robustness index of 5 control strategies is all undesirable.Can select the control accuracy of each control variable and the equal scaled down of the variation range of state variable, also in conjunction with the uncertainty of the controllability of each control variable with each state variable, the variation range of the control accuracy of each control variable and state variable can be carried out differentiation but not scaled down.For reducing the number of regulated variable, select that the latter is namely preferential regulates control variable that controllability is strong and the strong state variable of the large uncertainty of mobility scale.When robustness calculates, various discrete control variable to be fixed as in control strategy to be assessed corresponding value, below only each continuous type control variable controllability is calculated.
If the total number of control variable is M, i-th control variable X
irepresent, this control variable can control accuracy be D
i, it can span be Δ A in engineering
i, then its total quantization gear number is:
H
i＝ΔA
i/D
i
The coverage pattern supposing this control variable in the limited group of excellent solution assessed in pending robustness is Δ B
i, then the quantification gear number of its coverage pattern is:
ψ
i＝ΔB
i/D
i
The controllability subindex defining this control variable is S:
S＝1-ψ
i/H
i
Obviously, S is larger, and the ratio that this variable accounts for total span in excellent variation range of separating in set to be assessed is less, reflects that it is stronger to the controllability of performance index.
According to said method, continuous type control variable controllability result of calculation is as shown in table 9.
Table 9 each stepless control variable controllability calculates data and rank
Can be obtained by table 6, the descending rank of each continuous type control variable controllability is as follows: V3, V1, V4, V2, V6, V5.Can be obtained by table 3, three class state variable mobility scale (uncertainty) descending ranks are as follows: the 3rd class, the first kind, Equations of The Second Kind.According to control variable controllability and state variable uncertainty, the variable that will regulate is divided into three gradients: V1, V3 and the 3rd type load are the first gradient; V2, V4 and first kind load are the second gradient; V5, V6 and Equations of The Second Kind load are the 3rd gradient.Preferential adjustment affects the first large gradient to performance index, if result does not meet the demands, superposition regulates second and third gradient successively.
The control accuracy of each control variable in first gradient and the variation range of state variable are reduced into original 80%, robustness evaluates calculation result is as shown in table 7.
Can obtain with the Data Comparison of table 5 before adjustment, the robustness index of 5 groups of control strategies all raises, indicate the feasibility of the method, and robustness rank is consistent with before adjustment, also demonstrates the reliability of Reactive power control strategy being carried out to robustness appraisal procedure herein simultaneously.Can be obtained by table 10, the robustness index R of the 1st group of control strategy
cvalue is 0.8332, is greater than threshold value 0.83, demonstrates the quick validity by variable, Index Influence preferentially being regulated to first this method of gradient variable simultaneously.
Table 10 adjusts the robustness assessment result after relevant variable
Claims (4)
1. a robustness appraisal procedure for reactive power optimization of power system control strategy, is characterized in that, comprise the following steps:
1) be global space by combining by the span of each control variable and the variation range of each state variable the definition space formed in electric system to be optimized; By near the Reactive power control strategy C of certain pending robustness assessment, combine by the controlled accuracy rating of each control variable and the variation range of each state variable the space formed, be defined as robust space;
2) for single performance index, employing random sampling optimization method obtains the optimal performance index f in global space
b, the poorest performance index are f
w, and Reactive power control strategy C robust space to be assessed in optimal performance index f
b,c, the poorest performance index are f
w,c, corresponding minimum sampling number N
min, computing formula is:
For global space, the value of p% is 0.1% ~ 0.5%; The value of q% is 0.5% ~ 1%;
For robust space, the value of p% is 1% ~ 2%; The value of q% is 0.5% ~ 1%;
3) in global space, optimal performance index with the difference Δ f of the poorest performance index is:
Δf＝f
b-f
w
For Reactive power control strategy C to be assessed, in its robust space, the difference Δ f of optimal performance index and the poorest performance index
cfor:
Δf
c＝f
b,c-f
w,c
The robustness index R of definition control strategy C
cfor:
4) if R
c> threshold value R
set, then think that the robustness of Reactive power control strategy C meets the demands; Threshold value R
setbe more than or equal to 0.8.
2. the robustness appraisal procedure of reactive power optimization of power system control strategy according to claim 1, for the situation of multi-performance index, when only the robustness index of each performance index being undertaken assessing by step according to claim 1 and confirmed to meet the demands, just think that the robustness of Reactive power control strategy C meets the demands.
3. the robustness appraisal procedure of a kind of reactive power optimization of power system control strategy according to claim 1, it is characterized in that, several Reactive power control strategy is all met to the situation of robustness requirement, the mode taking equal proportion to expand the control accuracy of each control variable and the variation range of state variable gradually expands robust space, and re-start robustness assessment, do not meet the Reactive power control strategy of robustness requirement after the robustness first not meeting the Reactive power control strategy of robustness requirement is weaker than.
4. the robustness appraisal procedure of a kind of reactive power optimization of power system control strategy according to claim 1, it is characterized in that, several idle work optimization strategy is not all met to the situation of robustness requirement, by the control accuracy of each control variable and the equal scaled down of the variation range of state variable to β %, if the robustness within the scope of this meets the demands, then the receptance of the robustness of this idle work optimization strategy is claimed to be β %.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510461517.5A CN105005866A (en) | 2015-07-30 | 2015-07-30 | Robustness assessing method of power system reactive optimization control policy |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510461517.5A CN105005866A (en) | 2015-07-30 | 2015-07-30 | Robustness assessing method of power system reactive optimization control policy |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105005866A true CN105005866A (en) | 2015-10-28 |
Family
ID=54378529
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510461517.5A Pending CN105005866A (en) | 2015-07-30 | 2015-07-30 | Robustness assessing method of power system reactive optimization control policy |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105005866A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115021335A (en) * | 2022-06-16 | 2022-09-06 | 西安交通大学 | Multi-period robustness and reliability assessment method considering energy storage and new energy |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102170137A (en) * | 2011-04-26 | 2011-08-31 | 华北电力大学 | ORP (optimal reactive power) method of distribution network of electric power system |
| CN104037781A (en) * | 2014-05-21 | 2014-09-10 | 中国农业大学 | Reactive power grid coordination control method and device |
| CN104410077A (en) * | 2014-12-18 | 2015-03-11 | 积成电子股份有限公司 | Improved genetic algorithm based multithreading voltage and reactive power optimization control method for electric power system |
-
2015
- 2015-07-30 CN CN201510461517.5A patent/CN105005866A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102170137A (en) * | 2011-04-26 | 2011-08-31 | 华北电力大学 | ORP (optimal reactive power) method of distribution network of electric power system |
| CN104037781A (en) * | 2014-05-21 | 2014-09-10 | 中国农业大学 | Reactive power grid coordination control method and device |
| CN104410077A (en) * | 2014-12-18 | 2015-03-11 | 积成电子股份有限公司 | Improved genetic algorithm based multithreading voltage and reactive power optimization control method for electric power system |
Non-Patent Citations (5)
| Title |
|---|
| 吕文: "大规模风电并网无功优化控制策略", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 * |
| 孙兆键: "大规模风电场无功电压优化协调控制策略研究", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 * |
| 张勇: "变电站电压无功优化策略研究", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 * |
| 郭玉聪: "分散式风电接入配网无功优化控制", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 * |
| 陈卓等: "复杂电力系统鲁棒性协调控制研究", 《贵州工业大学学报(自然科学版)》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115021335A (en) * | 2022-06-16 | 2022-09-06 | 西安交通大学 | Multi-period robustness and reliability assessment method considering energy storage and new energy |
| CN115021335B (en) * | 2022-06-16 | 2024-03-26 | 西安交通大学 | Multi-period robust reliability assessment method considering energy storage and new energy |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Dobbe et al. | Toward distributed energy services: Decentralizing optimal power flow with machine learning | |
| Lim et al. | Robust decentralised load–frequency control of multi-area power systems | |
| Papamichail et al. | Traffic-responsive linked ramp-metering control | |
| Eqtami et al. | Event-triggered strategies for decentralized model predictive controllers | |
| CN112310980B (en) | Safety and stability evaluation method and system for direct-current blocking frequency of alternating-current and direct-current series-parallel power grid | |
| CN106959618A (en) | A kind of voltage control method for coordinating for optimizing weight based on ladder | |
| Zhang et al. | Real-time active and reactive power regulation in power systems with tap-changing transformers and controllable loads | |
| CN113285450A (en) | Power grid static voltage stabilization on-line prevention control optimization method | |
| Huang et al. | Fully decentralized multiarea reactive power optimization considering practical regulation constraints of devices | |
| CN113890017A (en) | Power distribution network voltage self-adaptive control method based on key measurement | |
| CN105005866A (en) | Robustness assessing method of power system reactive optimization control policy | |
| CN104155876B (en) | The separation of a kind of PID controller realizes method | |
| Luo et al. | Dynamic optimization of aeration operations for a benchmark wastewater treatment plant | |
| Beraud et al. | Optimization of WWTP control by means of multi-objective genetic algorithms and sensitivity analysis | |
| CN109301817A (en) | A kind of Multiple Time Scales source net lotus coordinated scheduling method considering demand response | |
| Ranamuka et al. | Mitigating tap changer limit cycles in modern electricity networks embedded with local generation units | |
| Binazadeh et al. | Passivity-based optimal control of discrete-time nonlinear systems | |
| Ortmann et al. | Online Feedback Optimization for Subtransmission Grid Control | |
| Tang et al. | Coordinated voltage control of weak sub-transmission networks considering wind power variability | |
| CN114944659B (en) | Automatic control method and system for wind-solar-energy-storage combined power station | |
| Wang et al. | Load frequency active disturbance rejection control for an interconnected power system via deep reinforcement learning | |
| Vrakopoulou et al. | Optimal control policies for reserve deployment with probabilistic performance guarantees | |
| Li et al. | Optimal coordinated voltage control of power systems | |
| Castro et al. | Comparative analysis among Pareto, Pareto—Fuzzy logic and Pareto—Fuzzy logic PI for distribution networks with the presence of DGs | |
| Khodabakhshian et al. | Design of a new load frequency PID controller using QFT |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20151028 |
|
| RJ01 | Rejection of invention patent application after publication |