CN102969720A - Load dynamic control and analysis method capable of being applied in smart power grids - Google Patents

Abstract

Disclosed is a load dynamic control and analysis method capable of being applied in smart power grids. The method comprises the steps of communication network analysis, wherein communication network performances of smart power grids are analyzed according to periodic transmission data of loads and combined with data acquired in real time to obtain performances of factors affecting performances of smart power grids at present; establishment of dynamic load analysis and control models, wherein corresponding dynamic load analysis and control models are established according to obtained load information and communication network performances, and load current data and stored historical data are analyzed to precast future power utilization conditions of loads; and load processing, wherein power allocation of loads are optimized and controlled based on the dynamic load analysis and control module predicted results.

Description

A kind of load that can use in intelligent grid is control and analytical method dynamically

Technical field

The present invention relates to a kind of load that can in intelligent grid, use dynamically control and analytical method, belong to communication and areas of information technology.

Background technology

Energy problem has become one of important topic of the world today.How to utilize more efficiently energy, how more effectively conserve energy has become the hottest point subject that world's numerous enterprises and scientific research organization are studied.In the recent period, many countries have all proposed package plan, encourage enterprise and scientific research organization larger energy to be dropped into research and the application in intelligent grid field.At present, electrical network has become basis and the important component part of industrialization, informationized society development.But under current expanding economy trend, traditional electrical network can not satisfy the demand of power industry, can't realize utilizing efficiently energy, and intelligent grid becomes the inevitable direction of future world power network development.Briefly, intelligent grid combines traditional electrical network, information and communication network, and three in one is worked in coordination, and promotes mutually, has realized high efficiency, reliability, the fail safe of electrical network service, has reduced the cost of operation and management, has reached energy-conservation purpose.Intelligent grid is that advanced person's sensing measurement technology, ICT (information and communication technology), analysis decision technology, automatic control technology and electricity power technology are combined, and has realized the new-modernization electrical network integrated with electrical network infrastructure height.

No matter inexorable trend and new trend as world's electrical network develops domestic or external, all give the attention of height and greatly concern to intelligent grid.In China, the development prospect of intelligent grid is very good.China State Grid Corporation of China has proposed extra-high voltage grid as a kind of key rack, develop in harmony between China's electrical networks at different levels, with this solid electrical network as the basis, take the information communication platform as support, take Based Intelligent Control as means, links is controlled and optimized, generating comprising electric power system, transmission of electricity, power transformation, distribution, electricity consumption and scheduling etc., utilize fully advanced control technology, information technology and the communication technology, structure is with interactive, digitlization, automation and information turn to the middle usefulness of having of principal character characteristic, autonomous innovation, intelligentized electrical network leading in the world, realize the height integral fusion of " flow of power; information flow; Business Stream ", realize the precision of energy supply, correspondenceization, mutualization and complementationization guarantee that the service of modernized electrical network reaches strong reliable, economical and efficient, clean environment firendly, transparent opening, friendly interactive target.Along with the continuous propelling of market-oriented reform, intelligent grid has become the only way of modern power network technical development.

As everyone knows, the development of intelligent grid and application have related to the correlation technique in a lot of fields, and wherein, one of technology of core is exactly information and the communication technology the most.2009, initiated by IEEE, and proposed Project 2030 plans, formulated a series of international standards about intelligent grid.At Project 2030 in the works, explicitly point out intelligent grid and comprise three parts: energy technologies, the communication technology and information technology.Therefore, communication and information technology are as the extremely important core technology of intelligent grid, also be the basis that intelligent grid is achieved simultaneously, how application message and the communication technology, realize the intellectuality of modern power network, utilize the real time information of electrical network, adjustment and the balance energy supply of intelligence have become the current study hotspot in the intelligent grid field of information and communication technology (ICT).

The communication in current intelligent grid field and information technology research focus mostly on the framework of Communications and Information Systems, actual intelligent grid does not control and analyzes model and the method for load in real time, therefore there is a kind of like this technical need, namely, need a kind of analysis and control method, can pass through communication network, load information in the current electrical network of real-time monitoring, by historical data and current data the electricity consumption situation of load is analyzed intelligently, and come the electricity consumption situation of control load by communication network.

Summary of the invention

The limitation of prior art in view of the above the invention discloses the method that a kind of load that can use is dynamically controlled and analyzed in intelligent grid, real-time control and analysis are carried out in the load in the intelligent grid, thereby reached energy-conservation purpose.

A kind of intelligent grid load based on wireless sensor network is dynamically controlled and analytical method, may further comprise the steps:

1) communication network analysis: according to the data of the cyclical transmission of the load data in conjunction with Real-time Collection, the COMMUNICATION NETWORK PERFORMANCES of intelligent grid is analyzed, obtained the current performance that affects the key element of intelligent grid performance;

2) dynamic load analysis and control model is set up: according to the performance of the information of obtaining load and communication network, set up corresponding load dynamically control and analytical model, the current data of load and the historical data of storing are analyzed, predicted the load electricity consumption situation in future;

3) load is processed: based on the result of described dynamic load analysis and control model prediction, the electricity consumption allotment of load is optimized control.

Described communication network analysis step comprises:

A) data acquisition and transmitting step, the data message of Real-time Collection intelligent grid load also is sent to Based Intelligent Control with the information that collects and analyzes the center;

B) COMMUNICATION NETWORK PERFORMANCES analytical procedure sends and the time delay of the communication network that desired rate causes, blocks and analyze data acquisition rate to the impact of whole COMMUNICATION NETWORK PERFORMANCES according to electric network data information, data, and then analyzes the performance of Current Communication Network network.

In the dynamic control of described load and the analytical model establishment step, according to the data message that obtains and the key element that affects intelligent grid efficient, set up dynamic control and analytical model to user load in the intelligent grid.

In the described load treatment step, information centre carries out certain prediction in conjunction with the real time data information of load to the power information in load future according to constructed dynamic control model result, and then the user power utilization situation in the intelligent grid analyzed and then electrical network conveying situation is carried out redistribution, so both can ensure active user's electricity consumption, can make again remaining electric weight can offer the user that other more need electricity consumption, thereby the electricity consumption situation of control load improves the efficient of load electricity consumption well.

In the described load treatment step, the key element of analyzing influence electric energy saving rate comprises change frequency (being the stability state of User Status) and network delay and the obstruction of the transmission rate of data acquisition frequency, data, the quantity of regional interior nodes (being the coverage of data center), node state.

Described dynamic load analysis and control model is as follows:

$C=\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{W}_i{S}_i---\left(1\right)$

0≤W _i≤1,0≤S _i≤1 (2)

Wherein, C represents this user's the consumption index for electric energy, S _iThe factor that affects the intelligent grid power consumption efficiency, W _iThe weight of corresponding influencing factor,

Concrete model is suc as formula shown in (15), (16), (17):

C=W _RS _R+W _VS _V+W _LS _L+W _PS _P (3)

0≤S _R,S _V,S _L,S _P≤1 (4)

0≤W _R,W _V,W _L,W _P≤1 (5)

W _R+W _V+W _L+W _P=1

(6)

Wherein,

W _R, W _V, W _L, W _PRepresent four kinds of each factor S that affect the intelligent grid power consumption efficiency _R, S _V, S _L, S _PShared importance in present analysis and forecast assessment system, the respective weights W of all factors _R, W _V, W _L, W _PSum equals 1;

S _RBe the excessive ratio of each user's load power consumption, be expressed as

W _rThe electric energy that does not have use, W _{Max_r}It is the maximum of user's electric energy affluence;

S _vBe the fluctuation situation of each user's load power consumption, be expressed as

T _AveThe time interval of the transformation between the user power utilization situation different conditions, T _{Max_v}Be expressed as the maximum time interval of the transformation between the user power utilization situation different conditions;

S _LBe the loss ratio of each user's Energy Transfer, be expressed as

D is the distance between the user of power plant and shortage of energy, d _MaxUltimate range;

S _PThe priority of user power utilization.

The calculation procedure of described network delay and blocking probability is:

For each node, time delay is T at ordinary times _D, representation node is attempted to send data to data and is controlled the time of information centre between receiving, and be T the service time of wireless sensor network _S, the stand-by period of data is T in the network _W, the triadic relation is as follows:

T _D=T _S+T _W (7)

Wherein,

T _S=T _L/T _V+T _C+T _P (8)

T _LThe length of packet, T _VThe transmission rate of data, T _CThe MAC time delay, T _PIt is the wireless transmission time delay.Wherein, MAC time delay, i.e. medium access control time delay, expression data be from obtaining intelligent electric meter allow to send, to the time that can send in the channel, and time of data acquisition channel namely;

T _CkThe MAC time delay on k the path of data from node i to node j, T _WkIt is the service time on k the path of data from node i to node j;

Therefore, the end-to-end overall time delay in the wireless sensor network is

$T_{\mathrm{Delay}\_\mathrm{overall}}=\underset{R}{\mathrm{\Σ}}({\mathrm{Delay}}_R\·{\mathrm{\γ}}_{i,j})=\underset{R}{\mathrm{\Σ}}(\underset{(\mathrm{ij},k)\∈R}{\mathrm{\Σ}}(T_{\mathrm{Ck}}+T_{\mathrm{Wk}}+T_L/T_V)\·{\mathrm{\γ}}_{\mathrm{ij},k})---\left(9\right)$

Wherein,

${\mathrm{Delay}}_R=\underset{i\∈R}{\mathrm{\Σ}}{D}_i---\left(10\right)$

Delay _RRepresentative is based on the overall average time delay of path R, γ _{I, j}The ratio set of representative data from node i to all paths of node j, γ _{Ij, k}The ratio of representative data from node i to k path of node j, both relations are as follows:

${\mathrm{\γ}}_{i,j}=\underset{(\mathrm{ij},k)\∈R}{\mathrm{\Σ}}{\mathrm{\γ}}_{\mathrm{ij},k}---\left(11\right)$

According to document (Yan Ye, Cai Hua, Seo Seung-Woo, Performance Analysisi of IEEE 802.11Wireless sensor networks[C], IEEE International Conference on Communications, pp.2547-2551,2008), the generating probability generating function of MAC time delay is as follows:

$T_{\mathrm{MAC}}\left(Z\right)=(1-p)S\left(Z\right)\underset{i=0}{\overset{L}{\mathrm{\Σ}}}\left\{{\left[\mathrm{PI}\left(Z\right)\right]}^i\underset{j=0}{\overset{i}{\mathrm{\Π}}}{D}_j\left(Z\right)\right\}+{\left[\mathrm{pI}\left(Z\right)\right]}^{L+1}\underset{i=0}{\overset{L}{\mathrm{\Π}}}{D}_i\left(Z\right)---\left(12\right)$

P refers to certain to other probability of communicating by letter and clashing in the communication of node and the collision domain, specifically is expressed as follows:

p=1-(1-τ) ^n-1 (13)

$S\left(Z\right)=Z^{T_S},I\left(Z\right)=Z^{T_I}\mathrm{and}{D}_i\left(Z\right)=\begin{array}{c}\left\{\begin{array}{c}\underset{i=0}{\overset{{\mathrm{CW}}_i-1}{\mathrm{\Σ}}}\frac{D\left(Z\right)}{{\mathrm{CW}}_i},0\≤i\≤m\\ D_m\left(Z\right),m\≤i\≤L\end{array}\right.---\left(14\right)\end{array}$

Obtain expectation and the variance of MAC time delay

(15)

E(T _MAC)=T′ _MAC(Z)| _Z=1

Var(T _MAC)=T″ _MAC(Z) _Z＝1+T′ _MAC(Z)| _Z=1-[T _MAC(Z)| _Z=1] ² (16)

Wherein,

L represents the number of times of maximum detection channel;

Independent variable Z is after the dispersed problem in the formation is carried out Z-transformation, problem is converted into become continuous problem on the Z territory so that carrying out mathematics solves, and what practical significance variable Z itself does not have;

Being calculated as follows of stand-by period:

At first, definition arrives the transmission density of load:

$\mathrm{\ρ}=\mathrm{\λb}=\frac{\mathrm{\λ}}{\mathrm{\μ}}---\left(17\right)$

Wherein, λ is the data arrival rates, and μ is the service speed of packet, and b is the packet desired value of service time;

If η is the probability of data-bag lost, P _kIt is the average probability that has k packet in the formation; E (X) and E (T) are par and the average delay of the data of service queue;

The relation of above parameter is as follows:

η＝λ(1-P _k) (18)

E(X)=rE(T) (19)

Pollaczek-Khinchin formula according in the document (Cassandras Christos G., Lafortune Stephane, " Introduction to Discrete Event Systems ", USA, 2009.) draws:

$E\left(X\right)=\frac{\mathrm{\ρ}}{1-\mathrm{\ρ}}-\frac{{\mathrm{\ρ}}^2}{2(1-\mathrm{\ρ})}(1-{\mathrm{\μ}}^2{\mathrm{\σ}}^2)---\left(20\right)$

Wherein, σ ²Be the variance of service time, X is the length of formation; Therefore, obtain the desired value E (T of service time _W):

$E\left(T_W\right)=E\left(T\right)-b=\frac{1}{\mathrm{\η}}E\left(X\right)-b---\left(21\right)$

Calculate the blocking probability in the whole network:

$P_{\mathrm{block}}=P_K=1-\frac{1}{{\mathrm{\π}}_0-\mathrm{\ρ}}---\left(22\right)$

Wherein, π ₀When the expression packet enters service queue, the initial condition of whole formation.

The flow chart of described method as shown in Figure 1.

The invention has the beneficial effects as follows, development trend for existing intelligent grid, provide a kind of available, efficient technical support and reference to intelligent grid in the innovation of the construction of Future Information communication network and energy-conservation prediction, analytical model so that intelligent grid can be better, grow up faster.The present invention can the combining information communication network performance and history and the present situation of electrical network load, carry out active data transmission, power delivery adjusting, the guarantee information communication network can be supported the operation of intelligent grid efficiently, the electricity consumption situation that guarantees each load in the intelligent grid obtains accurately, analysis and prediction targetedly, guarantee to load in following a period of time and utilize efficiently electric energy, put forward high-octane utilance.And for the situation that might cause a large amount of waste of energy, carry out to a certain degree prevention and adjusting, thereby reach the final purpose that intelligent grid is built---strong reliable, economical and efficient, clean environment firendly, transparent opening, friendly interactive modernized electrical network service are provided.

Description of drawings

Fig. 1 is the flow process general introduction figure based on the dynamic control of the intelligent grid load of wireless sensor network and analytical method;

Fig. 2 is the particular flow sheet based on the dynamic control of the intelligent grid load of wireless sensor network and analytical method;

Fig. 3 is the network topology schematic diagram of intelligent grid;

Fig. 4 is the change along with number of nodes, and communication network is the performance map of time delay end to end;

Fig. 5 is the change along with number of nodes, the performance map of communication network data generation blocking probability;

Fig. 6 is the change along with number of nodes, the performance comparison diagram of energy utilization efficiency;

Fig. 7 is the change along with data acquisition time, the performance comparison diagram of energy utilization efficiency;

Fig. 8 is that the change of number of nodes and data transmission rate is on the comparison diagram that affects of the utilization ratio of energy;

Fig. 9 is that the change of data acquisition rate and transmission rate is on the comparison diagram that affects of the utilization ratio of energy;

Figure 10 is that the change of data acquisition rate and number of nodes is to the accuracy of the demand control of load energy.

Embodiment

Performance based on the dynamic control of the load of wireless sensor network and analytical model mainly is to weigh by the efficient that energy is saved.The efficient that energy is saved is the ratio of energy that load the is saved energy of use when not using a model with load.If M is the energy that load is used when not using this dynamic control model, T is the energy that load is used when using this dynamic control model, and then the efficient of energy saving is Power Saving=M-TM.

For purpose of the present invention, feature and advantage can be become apparent more, the present invention is further detailed explanation below in conjunction with accompanying drawing.

Based on the intelligent grid load of wireless sensor network dynamically control and analytical method steps flow chart as shown in Figure 2. load information and the data of intelligent grid all gather by intelligent electric meter.The intelligent electric meter collection all be the real time data of load, the guarantee information center can obtain the current situation of load timely.Then, intelligent electric meter can periodically send data with certain speed to information centre.The cycle that data send and speed are that the situation according to current network determines that different cycles and speed can affect the performance of communication network, thereby affect the overall performance of intelligent grid.When load data is received by information centre, according to current information, analyze the performance of current network, determine parameters and the weight of the load control model, thereby obtain being applicable to load control and the analytical model of current intelligent grid.Information centre carries out analysis and prediction according to current model to the situation of intelligent grid, if the result shows that current load state need to carry out the redistribution of electric energy, will carry out redistribution to electric energy, proof load can more efficient the electric power, improves the efficiency of whole electrical network.Otherwise information centre can not redistribute electric energy.At last, information centre periodically carries out statistics and analysis to the energy-conservation situation of whole intelligent grid.

Fig. 3 is the network topology structure schematic diagram of the intelligent grid among the present invention.As shown in the figure, G _iI power plant, S _IjRepresent i j aggregation node under the power plant control, in this network, exist dissimilar users, have the power plant of some, each power plant provides electric energy for the user of some simultaneously, and each power plant is directly to link to each other.In intelligent grid, exist intelligent electric meter, it can be real-time monitor user ' information about power, collect data and data message be sent to Based Intelligent Control and the information centre in power plant by wireless sensor network, Based Intelligent Control and information centre analyze model prediction user's following electricity consumption situation by Based Intelligent Control, thereby can take full advantage of by the electric energy that a part is superfluous.The user's in part zone power information can be sent to by this aggregation node Based Intelligent Control and the information centre in power plant.

Fig. 4, Fig. 5 are to the analytical performance figure of the communication network of intelligent grid among the present invention.From figure, can find that user's quantity increases, the speed difference of intelligent electric meter transmission data can change end-to-end time delay and the blocking probability of communication network.Make a concrete analysis of as follows:

For each node, average delay is T _D, be T the service time of wireless sensor network _S, the stand-by period of data is T in the network _W, the triadic relation is as follows:

T _D=T _S+T _W (7)

Wherein,

T _S=T _L/T _V+T _C+T _P (8)

T _LThe length of packet, T _VThe transmission rate of data, T _CThe MAC time delay, T _PIt is the wireless transmission time delay.

T _CkThe MAC time delay on k the path of data from node i to node j, T _WkIt is the service time on k the path of data from node i to node j;

Therefore, the end-to-end overall time delay in the wireless sensor network is

$T_{\mathrm{Delay}\_\mathrm{overall}}=\underset{R}{\mathrm{\Σ}}({\mathrm{Delay}}_R\·{\mathrm{\γ}}_{i,j})=\underset{R}{\mathrm{\Σ}}(\underset{(\mathrm{ij},k)\∈R}{\mathrm{\Σ}}(T_{\mathrm{Ck}}+T_{\mathrm{Wk}}+T_L/T_V)\·{\mathrm{\γ}}_{\mathrm{ij},k})---\left(9\right)$

Wherein,

${\mathrm{Delay}}_R=\underset{i\∈R}{\mathrm{\Σ}}{D}_i---\left(10\right)$

Delay _RRepresentative is based on the overall average time delay of path R, γ _{I, j}The ratio set of representative data from node i to all paths of node j, γ _{Ij, k}The ratio of representative data from node i to k path of node j, both relations are as follows:

${\mathrm{\γ}}_{i,j}=\underset{(\mathrm{ij},k)\∈R}{\mathrm{\Σ}}{\mathrm{\γ}}_{\mathrm{ij},k}---\left(11\right)$

The generation probability function of MAC time delay is as follows:

$T_{\mathrm{MAC}}\left(Z\right)=(1-p)S\left(Z\right)\underset{i=0}{\overset{L}{\mathrm{\Σ}}}\left\{{\left[\mathrm{pI}\left(Z\right)\right]}^i\underset{j=0}{\overset{i}{\mathrm{\Π}}}{D}_j\left(Z\right)\right\}+{\left[\mathrm{pI}\left(Z\right)\right]}^{L+1}\underset{i=0}{\overset{L}{\mathrm{\Π}}}{D}_i\left(Z\right)---\left(12\right)$

P refers to certain to other probability of communicating by letter and clashing in the communication of node and the collision domain, specifically is expressed as follows:

p=1-(1-τ) ^n-1 (13)

$S\left(Z\right)=Z^{T_S},I\left(Z\right)=Z^{T_I}\mathrm{and}{D}_i\left(Z\right)=\left\{\begin{array}{c}\underset{i=0}{\overset{{\mathrm{CW}}_i-1}{\mathrm{\Σ}}}\frac{D\left(Z\right)}{{\mathrm{CW}}_i},0\≤i\≤m\\ D_m\left(Z\right),m\≤i\≤L\end{array}\right.---\left(14\right)$

Obtain expectation and the variance of MAC time delay

(15)

E(T _MAC)=T′ _MAC(Z)| _Z=1

Var(T _MAC)=T″ _MAC(Z)| _Z＝1+T′ _MAC(Z)| _Z=1-[T _MAC(Z)| _Z=1] ² (16)

Wherein,

L represents the number of times of maximum detection channel;

Independent variable Z is after the dispersed problem in the formation is carried out Z-transformation, problem is converted into become continuous problem on the Z territory so that carrying out mathematics solves, and what practical significance variable Z itself does not have;

Being calculated as follows of stand-by period:

At first, definition arrives the transmission density of load:

$\mathrm{\ρ}=\mathrm{\λb}=\frac{\mathrm{\λ}}{\mathrm{\μ}}---\left(17\right)$

Wherein, λ is the data arrival rates, and μ is the service speed of packet, and b is the packet desired value of service time;

If η is the probability of data-bag lost, P _kIt is the average probability that has k packet in the formation; E (X) and E (T) are par and the average delay of the data of service queue;

The relation of above parameter is as follows:

η＝λ(1-P _k) (18)

E(X)=ηE(T) (19)

$E\left(X\right)=\frac{\mathrm{\ρ}}{1-\mathrm{\ρ}}-\frac{{\mathrm{\ρ}}^2}{2(1-\mathrm{\ρ})}(1-{\mathrm{\μ}}^2{\mathrm{\σ}}^2)---\left(20\right)$

Wherein, σ ²Be the variance of service time, X is the length of formation; Therefore, obtain the desired value E (T of service time _W):

$E\left(T_W\right)=E\left(T\right)-b=\frac{1}{\mathrm{\η}}E\left(X\right)-b---\left(21\right)$

Calculate the blocking probability in the whole network:

$P_{\mathrm{block}}=P_K=1-\frac{1}{{\mathrm{\π}}_0-\mathrm{\ρ}}---\left(22\right)$

Wherein, π ₀When the expression packet enters service queue, the initial condition of whole formation.

By above analysis, we can obtain time delay in the wireless sensor network and the computational methods of blocking probability, thereby analyze the important performance of whole network.Such as Fig. 4, shown in Figure 5, significantly increasing can appear in the time delay in the network and block raising along with the increasing of user, transmission rate.Therefore, in the middle of the Network Topology Design of intelligent grid, can not each regional number of users of unconfined increase, but to reasonably distribute so that communication network is in the middle of the optimized state, so that the performance of intelligent grid can better be supported and promote to communication network, guarantee efficient, the operation at a high speed of intelligent grid, improve energy-efficiency, reach energy-conservation effect.

The load control model that intelligent grid load control and information centre adopt is considerable in the present invention.Dynamic load analysis and control model is as follows:

$C=\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{W}_i{S}_i---\left(15\right)$

0≤W _i≤1，0≤S _i≤1 (16)

Wherein, S _iThe factor that affects the intelligent grid power consumption efficiency, W _iIt is the weight of corresponding influencing factor.

In the present invention, because experimental situation is limited, only considered four important factors, concrete model is suc as formula shown in (15), (16), (17):

C=W _RS _R+W _VS _V+W _LS _L+W _PS _P

(17)

0≤S _R,S _V,S _L,S _P≤1

(18)

0≤W _R,W _V,W _L,W _P≤1 (19)

Wherein,

1.S _RBe the excessive ratio of each user's load power consumption, be expressed as

W _rThe electric energy that does not have use, W _{Max_r}It is the maximum of user's electric energy affluence.

2.S _vBe the fluctuation situation of each user's load power consumption, be expressed as

T _AveThe time interval of the transformation between the user power utilization situation different conditions, T _{Max_v}Be expressed as the maximum time interval of the transformation between the user power utilization situation different conditions.

3.S _LBe the loss ratio of each user's Energy Transfer, be expressed as

D is the distance between the user of power plant and shortage of energy, d _MaxUltimate range.

4.S _PThe priority of user power utilization.

Fig. 6 is the change along with number of nodes, the performance comparison diagram of the energy utilization efficiency of whole intelligent grid.Wherein, W _R=W _V=W _L=W _P=0.25, the shared importance ratio of various key elements is identical.Number of users among the figure in the whole network is incremented to 1000 gradually from 100, can find out, along with increasing of user node quantity, the utilization ratio of energy is constantly increasing, this mainly is because user node is more, the electric energy operating position can be complementary user node more, remaining electric energy also can be utilized more fully.But when user node quantity arrived the quantity of some, the capacity usage ratio in the whole network be there will be no larger increase, and comparatively stably situation occurred.In Fig. 6, the time of having adopted different User Status to change is respectively 0.5 hour, 1 hour, 5 hours, 10 hours, 15 hours simultaneously.According to curve among the figure, the average time that the User Status of can reaching a conclusion changes is longer, and the utilance of energy is higher, can reach ten Percent five.This illustrates that also the user's is more stable with electricity condition, and namely User Status change frequency is lower, more is conducive to put forward high-octane utilization ratio.

Fig. 7 is the change along with data acquisition time, the performance comparison diagram of energy utilization efficiency.Can be found out that by curve among the figure user's state is more steady, stable, it is larger that state changes the time interval, and the time interval of image data is just larger, and the data of communication network transmission are just fewer, certainly just more favourable to the raising of the capacity usage ratio of whole intelligent grid.When user's the state change time surpassed certain numerical value, the capacity usage ratio of whole intelligent grid tended to be steady.

Fig. 8 is that the change of user node quantity and data transmission rate is on the comparison diagram that affects of the utilization ratio of energy.3 curves represent that user node transmission rate λ is 10,30 and 60 o'clock energy saving rate among the figure.By contrasting the ratio of the energy saving under different transmission rates and the desirable transmission rate state, obtain as drawing a conclusion: in the time of the user node negligible amounts, user node uses different rate sending datas, do not cause performance larger difference to occur, all the performance with perfect condition is suitable.

Can find out by Fig. 4, Fig. 5, user node quantity is few, transmission rate can't cause the performance of network larger fluctuation to occur, does not also affect the transmission of the real time data of communication network in the intelligent grid, since also just can not affect the overall performance of the electric energy service efficiency of intelligent grid.But along with the continuous increase of user node quantity, user node uses different rate sending datas, and overall performance will larger difference occur.The increase of network delay, obstruction, conflict will larger impact occur to the transmission of real time data.The speed that data send is larger, the performance of wireless sensor network will be poorer, cause the information control center that is sent to that data can not be real-time, thereby the state that causes users, such as dump energy or lack electric weight, can't carry out timely balance by information control center, therefore will cause the capacity usage ratio of intelligent grid lower.

Fig. 9 is that the change of data acquisition rate and transmission rate is on the comparison diagram that affects of the utilization ratio of energy.Among this figure so that the speed that the factor of difference mainly is change, data transmission rate and the data acquisition of User Status appears in overall performance.User's speed more stable with electricity condition, data acquisition is just less, needs the electricity consumption data of transmission just fewer, and also less with regard to impact on the performance of integral body, in addition, data transmission rate is less also can to produce less impact.

Figure 10 is that the change of data transmission rate and number of nodes is to the accuracy of the demand control of load energy.This figure main manifestations be the data that different data transmission rate and user node quantity are sent to information control center, and then according to analysis result, the result that user's electricity consumption situation is predicted and controlled.Near might distributing to the remaining electric weight of some user in this process the user who lacks electric weight, the result who has realized saving electric weight.Also might send electric energy to do not need electric energy user, error occur.Therefore, information control center is quite important to the control rate of precision of user load electric energy.Can find out obviously that from figure the quantity of user node is larger, data transmission rate is larger, and the accuracy of control is poorer.

By the above concrete analysis that how different factors is affected the energy saving rate, can draw to draw a conclusion: under the perfect condition, data acquiring frequency is lower, transmission rate is less, data jamming with conflict littlely, the user is less, User Status is more stable, network delay is less, the overall performance of intelligent grid is better.But in actual applications, must suit measures to local conditions, user in the intelligent grid is carried out suitable Regional Differentiation, the different periods are adopted different data acquisition, transmission rate, adapt with user's state, the accuracy that user's electric weight distributes, this in fact also is the process of a game, so that the overall performance of intelligent grid reaches optimization.Certainly, along with the development of intelligent grid, the factor that needs to consider will get more and more, and the parameter that the load dynamic control model is considered also will get more and more.

Claims (7)

1. an intelligent grid load is dynamically controlled and analytical method, it is characterized in that, may further comprise the steps:

Communication network analysis: according to the data of the cyclical transmission of the load data in conjunction with Real-time Collection, the COMMUNICATION NETWORK PERFORMANCES of intelligent grid is analyzed, obtained the current performance that affects the key element of intelligent grid performance;

Dynamic load analysis and control model is set up: according to the performance of the information of obtaining load and communication network, set up corresponding dynamic load analysis and control model, the current data of load and the historical data of storing are analyzed, predicted the load electricity consumption situation in future;

Load is processed: based on the result of described dynamic load analysis and control model prediction, the electricity consumption allotment of load is optimized control.

2. a kind of intelligent grid load according to claim 1 is dynamically controlled and analytical method, it is characterized in that described COMMUNICATION NETWORK PERFORMANCES analytical procedure comprises:

Data acquisition and transmitting step, the data message of Real-time Collection intelligent grid load also is sent to Based Intelligent Control with the information that collects and analyzes the center;

The COMMUNICATION NETWORK PERFORMANCES analytical procedure sends and the time delay of the communication network that desired rate causes, blocks and analyze data acquisition rate to the impact of whole COMMUNICATION NETWORK PERFORMANCES according to electric network data information, data, and then analyzes the performance of Current Communication Network network.

3. a kind of intelligent grid load according to claim 1 is dynamically controlled and analytical method, it is characterized in that, in the described dynamic load analysis and control model establishment step, data message according to above-mentioned gained, according to the key element that affects intelligent grid efficient, set up the dynamic load analysis and control model to user load in the intelligent grid.

4. a kind of intelligent grid load according to claim 1 is dynamically controlled and analytical method, it is characterized in that, in the described load treatment step, information centre predicts the power information in load future in conjunction with the real time data information of load according to constructed dynamic control model result, and then the user power utilization situation in the intelligent grid is analyzed and then electrical network conveying situation is carried out redistribution.

5. a kind of intelligent grid load according to claim 1 is dynamically controlled and analytical method, it is characterized in that, in the described load treatment step, the key element of analyzing influence electric energy saving rate comprises change frequency (being the stability state of User Status) and network delay and the obstruction of the transmission rate of data acquisition frequency, data, the quantity of regional interior nodes (being the coverage of data center), node state.

6. according to claim 1,2,3,4 or 5 described a kind of intelligent grid loads dynamically control and analytical method, it is characterized in that dynamic load analysis and control model is as follows:

$C=\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{W}_i{S}_i---\left(1\right)$

0≤W _i≤1,0≤S _i≤1 (2)

Wherein, C represents this user's the consumption index for electric energy, S _iThe factor that affects the intelligent grid power consumption efficiency, W _iThe weight of corresponding influencing factor,

Concrete model is suc as formula shown in (15), (16), (17):

C=W _RS _R+W _VS _V+W _LS _L+W _PS _P (3)

0≤S _R,S _V,S _L,S _P≤1 (4)

0≤W _R,W _V,W _L,W _P≤1 (5)

W _R+W _V+W _L+W _P=1

(6)

Wherein,

W _R, W _V, W _L, W _PRepresent four kinds of each factor S that affect the intelligent grid power consumption efficiency _R, S _V, S _L, S _PShared importance in present analysis and forecast assessment system, the respective weights W of all factors _R, W _V, W _L, W _PSum equals 1;

S _RBe the excessive ratio of each user's load power consumption, be expressed as

W _rThe electric energy that does not have use, W _{Max_r}It is the maximum of user's electric energy affluence;

S _vBe the fluctuation situation of each user's load power consumption, be expressed as

T _AveThe time interval of the transformation between the user power utilization situation different conditions, T _{Max_v}Be expressed as the maximum time interval of the transformation between the user power utilization situation different conditions;

S _LBe the loss ratio of each user's Energy Transfer, be expressed as

D is the distance between the user of power plant and shortage of energy, d _MaxUltimate range;

S _PThe priority of user power utilization.

7. a kind of intelligent grid load according to claim 5 is dynamically controlled and analytical method, it is characterized in that the calculation procedure of network delay and blocking probability:

For each node, time delay is T at ordinary times _D, representation node is attempted to send data to data and is controlled the time of information centre between receiving, and be T the service time of wireless sensor network _S, the stand-by period of data is T in the network _W, the triadic relation is as follows:

T _D=T _S+T _W (7)

Wherein,

T _S=T _L/T _V+T _C+T _P (8)

T _LThe length of packet, T _VThe transmission rate of data, T _CThe MAC time delay, T _PIt is the wireless transmission time delay.Wherein, MAC time delay, i.e. medium access control time delay, expression data be from obtaining intelligent electric meter allow to send, to the time that can send in the channel, and time of data acquisition channel namely;

T _CkThe MAC time delay on k the path of data from node i to node j, T _WkIt is the service time on k the path of data from node i to node j;

Therefore, the end-to-end overall time delay in the wireless sensor network is

$T_{\mathrm{Delay}\_\mathrm{overall}}=\underset{R}{\mathrm{\Σ}}({\mathrm{Delay}}_R\·{\mathrm{\γ}}_{i,j})=\underset{R}{\mathrm{\Σ}}(\underset{(\mathrm{ij},k)\∈R}{\mathrm{\Σ}}(T_{\mathrm{Ck}}+T_{\mathrm{Wk}}+T_L/T_V)\·{\mathrm{\γ}}_{\mathrm{ij},k})$

(9)

Wherein,

${\mathrm{Delay}}_R=\underset{i\∈R}{\mathrm{\Σ}}{D}_i---\left(10\right)$

Delay _RRepresentative is based on the overall average time delay of path R, γ _{I, j}The ratio set of representative data from node i to all paths of node j, γ _{Ij, k}The ratio of representative data from node i to k path of node j, both relations are as follows:

${\mathrm{\γ}}_{i,j}=\underset{(\mathrm{ij},k)\∈R}{\mathrm{\Σ}}{\mathrm{\γ}}_{\mathrm{ij},k}---\left(11\right)$

The generation probability function of MAC time delay is as follows:

$T_{\mathrm{MAC}}\left(Z\right)=(1-p)S\left(Z\right)\underset{i=0}{\overset{L}{\mathrm{\Σ}}}\left\{{\left[\mathrm{PI}\left(Z\right)\right]}^i\underset{j=0}{\overset{i}{\mathrm{\Π}}}{D}_j\left(Z\right)\right\}+{\left[\mathrm{pI}\left(Z\right)\right]}^{L+1}\underset{i=0}{\overset{L}{\mathrm{\Π}}}{D}_i\left(Z\right)---\left(12\right)$

P refers to certain to other probability of communicating by letter and clashing in the communication of node and the collision domain, specifically is expressed as follows:

p=1-(1-τ) ^n-1 (13)

$S\left(Z\right)=Z^{T_S},I\left(Z\right)=Z^{T_I}\mathrm{and}{D}_i\left(Z\right)=\begin{array}{c}\left\{\begin{array}{c}\underset{i=0}{\overset{{\mathrm{CW}}_i-1}{\mathrm{\Σ}}}\frac{D\left(Z\right)}{{\mathrm{CW}}_i},0\≤i\≤m\\ D_m\left(Z\right),m\≤i\≤L\end{array}\right.---\left(14\right)\end{array}$

Obtain expectation and the variance of MAC time delay

(15)

E(T _MAC)=T′ _MAC(Z)| _Z=1

Var(T _MAC)=T″ _MAC(Z)| _Z＝1+T′ _MAC(Z)| _Z＝1-[T _MAC(Z)| _Z=1] ² (16)

Wherein,

L represents the number of times of maximum detection channel;

Independent variable Z is after the dispersed problem in the formation is carried out Z-transformation, problem is converted into become continuous problem on the Z territory so that carrying out mathematics solves, and what practical significance variable Z itself does not have;

Being calculated as follows of stand-by period:

At first, definition arrives the transmission density of load:

$\mathrm{\ρ}=\mathrm{\λb}=\frac{\mathrm{\λ}}{\mathrm{\μ}}---\left(17\right)$

Wherein, λ is the data arrival rates, and μ is the service speed of packet, and b is the packet desired value of service time;

If η is the probability of data-bag lost, P _kIt is the average probability that has k packet in the formation; E (X) and E (T) are par and the average delay of the data of service queue;

The relation of above parameter is as follows:

η＝λ(1-P _k) (18)

E(X)=ηE(T) (19)

$E\left(X\right)=\frac{\mathrm{\ρ}}{1-\mathrm{\ρ}}-\frac{{\mathrm{\ρ}}^2}{2(1-\mathrm{\ρ})}(1-{\mathrm{\μ}}^2{\mathrm{\σ}}^2)---\left(20\right)$

Wherein, σ ²Be the variance of service time, X is the length of formation; Therefore, obtain the desired value E (T of service time _w):

$E\left(T_W\right)=E\left(T\right)-b=\frac{1}{\mathrm{\η}}E\left(X\right)-b---\left(21\right)$

Calculate the blocking probability in the whole network:

$P_{\mathrm{block}}=P_K=1-\frac{1}{{\mathrm{\π}}_0-\mathrm{\ρ}}---\left(22\right)$

Wherein, π ₀When the expression packet enters service queue, the initial condition of whole formation.

Images