CN115079571B - Intelligent optimization control method for satellite power supply fault reconstruction based on wild goat algorithm - Google Patents
Intelligent optimization control method for satellite power supply fault reconstruction based on wild goat algorithm Download PDFInfo
- Publication number
- CN115079571B CN115079571B CN202210735650.5A CN202210735650A CN115079571B CN 115079571 B CN115079571 B CN 115079571B CN 202210735650 A CN202210735650 A CN 202210735650A CN 115079571 B CN115079571 B CN 115079571B
- Authority
- CN
- China
- Prior art keywords
- fault
- load
- group
- reconstruction
- weight
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/04—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators
- G05B13/042—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators in which a parameter or coefficient is automatically adjusted to optimise the performance
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/04—Constraint-based CAD
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
- Y04S10/52—Outage or fault management, e.g. fault detection or location
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Theoretical Computer Science (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Automation & Control Theory (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention discloses an intelligent optimization control method for satellite power supply fault reconstruction based on a wild goat algorithm, and relates to the technical field of power electronics. According to the invention, a multi-target function of maximum recovery load and minimum power loss of a load grade under the condition of a fault is established by sampling the electric quantity through a sensor, power construction and the like are determined as constraint conditions, and the weight occupied by two targets in the multi-target function is determined by applying a linear weighting method, so that a comprehensive target model of a system is constructed to obtain an optimal fault recovery switch sequence. According to the method, the wild goat algorithm and the intelligent control unit are combined, the wild goat algorithm is introduced on the basis of considering multi-objective function optimization, the speed of solving the optimal solution of the reconstruction algorithm can be effectively increased, and the convergence speed is better. The use of combining intelligent control unit can realize the double detection of reconstruction scheme, prevent to take place the secondary failure, provide certain fault-tolerant rate for system recovery, improved the accuracy of reconstruction result.
Description
Technical Field
The invention relates to the technical field of power electronics, in particular to an intelligent optimization control method for satellite power supply fault reconstruction based on a wild goat algorithm.
Background
Because the complexity of the spacecraft task is higher and higher, the power of load equipment is increased continuously, and the requirement on reliability is gradually improved, the satellite power supply and distribution system gradually develops towards the direction of fault autonomous diagnosis and autonomous recovery. The current more common method for realizing fault reconstruction consists of a ground detection system and an improved reconstruction algorithm. However, the implementation of the reconstruction scheme based on the ground detection system depends on data transmission of the satellite computer and the ground communication detection system, and after consultation by a ground expert, a reconstruction instruction is sent to the satellite for fault reconstruction, and this control method causes waste of fault reconstruction time, poor real-time detection performance, and risk of fault expansion. The most common algorithm for power restoration can realize real-time detection of faults and autonomous restoration on the satellite, but the wide-area solving algorithm has low solving efficiency and is easy to fall into a local optimal solution, so that the solving accuracy is not high, and in this respect, the reconstruction method provided for the dynamic purpose has a weak optimization result and low precision.
Disclosure of Invention
The invention aims to solve the technical problems that an intelligent optimization control method for satellite power supply fault reconstruction based on a wild goat algorithm is provided, the real-time detection performance of fault reconstruction is poor, the efficiency of a wide-area solving algorithm is low, and the current situation of local optimization is easy to fall into. The method of the invention can detect the fault in real time and improve the accuracy and rapidity of the reconstruction scheme.
In order to solve the technical problems, the technical scheme adopted by the invention is as follows: the intelligent optimization control method for satellite power supply fault reconstruction based on the wild goat algorithm comprises the following steps:
the method comprises the following steps that S1, a hierarchical control system is designed, wherein the hierarchical control system comprises a mathematical physical layer and a network information layer which performs information interaction with the mathematical physical layer through a sensor, the mathematical physical layer comprises a sampling unit and a control unit, and the network information layer comprises a fault detection unit, a fault area detection unit, a fault reconstruction unit and an intelligent control agent unit; the sensor transmits physical quantity obtained by sampling the load output end by the mathematical physical layer to the network information layer for data detection;
s2, detecting sampling parameters, judging whether a fault occurs or not, if the signal abnormality is detected, judging that the fault occurs, positioning the fault, and respectively transmitting fault information to the S3 and the S4; if the signal is detected to be normal, the step S5 is carried out, a normal signal is sent to the intelligent control agent, and the intelligent control agent continues to work;
s3, after receiving the fault signal and the fault positioning information, searching the optimal switch sequence of the system to reconstruct the fault based on the wild goat algorithm under the premise of considering the load grade, obtaining a reconstruction scheme of optimal power restoration under the load grade, and transmitting the reconstruction scheme to the intelligent control system in the step S5 to switch the switch;
s4, when the fault signal is received, the fault area detection unit positions and calculates the total load capacity of the fault area detected by the fault information
And will load the total capacity
Transmitting to an intelligent control unit to provide line capacity constraint for the intelligent control system in the step S5;
step S5, after receiving the signals from the system, the intelligent control unit sends switching instructions to the lines of each load, and in a normal mode of the system: after receiving the normal signal of the step S2, the system does not adjust the switching sequence and keeps the current switching mode; in a system failure mode: after the fault reconstruction sequence of the step S3 is received, the operation of the step S4 is carried out to switch the optimal switching sequence sent by the line according to the load priority and the line capacity;
and S6, after the fault reconstruction is completed, detecting a current electric quantity signal, transmitting the current electric quantity signal to a physical layer through a sensor to perform a new round of closed-loop control on the electric quantity of the line, and repeating the actions.
The technical scheme of the invention is further improved as follows: in the step S1, the physical quantity obtained by sampling the load output end by the mathematical physical layer includes the priority ω of the N loads i Resistance R on the ith branch in different switch conducting states i Current I i Voltage V i Load L on ith bus i And load state y i Load state y i The normal state is represented as 1 and the failure state is represented as 0.
The technical scheme of the invention is further improved as follows: the process of detecting the sampling parameters in the step S2 is as follows: according to the branch current I i Voltage V i And whether the values of the bus current and the bus voltage exceed the threshold values thereof, namely: i is i ＞I imax And V i ＞V imax (ii) a When the detection unit detects that the electric quantity values of the system are normal, the detection unit judges that the line works in a normal state, and the detection unit sends a normal signal to the intelligent control unit; when the real-time detection signal is abnormal, the detection unit judges that the line has a fault, and respectively transmits an abnormal signal to the fault reconstruction unit in the step S3 to enable the fault reconstruction unit to execute a reconstruction function; failure information is transmitted to the failure region detection in step S4 to calculate the failure capacity.
The technical scheme of the invention is further improved as follows: the specific steps of searching the optimal switch sequence of the system for fault reconstruction based on the wild goat algorithm under the premise of considering the load grade in the step S3 are as follows:
step S31, composition of multi-objective function model
Using linear weighting according to the maximum recovery load f 1 And minimum power loss f 2 The importance degree of the target function is used as weight, and a comprehensive objective function F is obtained through calculation, wherein the formula is as follows:
F＝λ 1 f 1 +λ 2 f 2
wherein, ω is i Is the priority, R, of the load i Is the resistance, L, on the ith branch in different switch on states i Is the load on the ith bus, y i Is a load state, a load state y i The normal state is represented as 1, and the fault state is represented as 0, lambda 1 And λ 2 Is a constant;
step S32, constraint conditions of inequality: the inequality constraint conditions comprise node voltage limit value constraint, line transmission power constraint and transmission line current constraint;
1) Node voltage limit constraints
After the satellite power supply is connected to the power distribution network, the voltage V of each node i The relevant constraints are satisfied as follows:
V imax ＞V i ＞V imin ；
2) Line transmission power constraints
The transmission power of the line between the nodes meets the relevant constraint conditions as follows:
P ij ＜P ijmax ，
P ij representing the power, P, of the line transmission between nodes ijmax Is the maximum value of the transmission power of the line;
3) Transmission line current constraints
Current I flowing in the line i The relevant constraints are satisfied as follows:
I imax ＞I i ＞I imin ；
step S33, initializing population: the load switches are numbered, binary coding is adopted, the switch states are represented by 0 and 1, the switch-off state is 0, the switch-on state is 1, and the switch states are combined together to form a problem solution omega g i Wherein ω g i ＝f[x i1 ,x i2 ,...,x iN ]；
Step S34, evaluating the individuals: the weight of each independent solution is calculated according to the comprehensive objective function:
wherein N is var Is the dimension of each problem optimization, N wg Is the population membership, f (wg) i ) Is wg per individual i Weight value of (2), F (wg) i ) Is the value of the integrated objective function;
then, grouping the individuals with equal numbers, and calculating the proportion of each individual in the group, wherein the weight of the optimal switching sequence is 1, and the others are 0-1, the individual with the high weight of the group is selected as a leader of the group, and the other individuals are followers;
step S35, group movement: after grouping, each group will move towards the best try direction, and the leader and follower in the group will have two different movement trends, as follows:
1) The leader will follow its own motion vector v i (t) and best try direction p i (t) carrying out group movement with followers in the collar group;
2) Direction v of movement of follower ik (t) there are three trends of influence: the follower's own motion vector and best try direction and the leader's motion direction;
the following is the calculation of the motion formulas for the leader and follower, respectively:
the leader: v i (t+1)＝ω×v i (t)+R×rand×(p i (t)-ωg i (t))，
Following the person: v ik (t+1)＝ω×v i (t)+R×rand×(p i (t)-ωg i (t))+W lk (t)×rand×(ωg lk (t)-ωg i (t))，
Where ω represents the inertial weight, R represents the learning coefficient, W lk Is the weight of the kth leader, and rand is a random parameter, typically taken as 0 or 1,v i (t) is the motion vector at time t, p i (t) is the best try direction, [ omega ] g i (t) is the ith individual weight, [ omega ] g lk (t) is the average weight of all members of the population at time t, v ik (t) is the motion vector of the follower;
step S36, reevaluation: re-evaluating each group in each population after mobile updating, and re-calculating the weight of each independent individual and selecting a leader;
step S37, group cooperation: in the process of searching for the optimal switching sequence, the weights of individuals in groups are updated regularly in the population, leader weights are compared among the populations, the group with high weight attracts the individual with the lowest weight in the lower group, and then the weight of the group is reduced, and the weight of the population is calculated as follows:
with increasing number of iterations, the group with low leader weight gradually loses its members of the group, even the leader, which is the best sequence for reconstruction when different groups of independent individuals merge into the same group due to mutual attraction of weights and possess only one leader.
The technical scheme of the invention is further improved as follows: in the system failure mode in step S5, the intelligent control unit performs the following operations: calculating the available power capacity ARC of the circuit shown by the switching sequence after the optimal switching sequence of the system is obtained according to the step S3 i Then the ARC is applied i And total capacity of fault load
Comparing, and judging whether the current line can support system recovery;
if it is satisfied with
Then the normal recovery of the system is carried out; if the load is not satisfied, the priority load grade is high, the recovery is carried out, then a switch exchange table is established, the switch capacity of the nearby line is calculated, the pressure of the load recovery is transmitted to the nearby line, and when the capacity of each line meets the requirement, but partial loads with low grades are not recovered by power supply, the off-load reconstruction is carried out.
Due to the adoption of the technical scheme, the invention has the technical progress that:
1. the invention can detect the working state of the power supply circuit in real time, and reduces the danger of fault expansion;
2. according to the solving scheme based on the wild goat algorithm, on the basis of the multi-objective optimization function, the solving speed of the reconstruction sequence can be effectively increased, the situation that the reconstruction sequence is trapped in local optimization is avoided, and the accuracy of system solving is improved;
3. the invention provides a method for combining intelligent control units, which can realize double detection of a reconstruction scheme on the basis of obtaining an optimal reconstruction sequence of a system, provide a certain fault-tolerant rate for system recovery and improve the accuracy of a reconstruction result.
Drawings
FIG. 1 is an overall block diagram of the operation of the present invention;
FIG. 2 is a schematic diagram of the power switch in step S33 according to the present invention;
fig. 3 is a schematic diagram of the operation of the intelligent control unit of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the following examples:
the following provides a detailed description of the preferred embodiments of the present invention with reference to the accompanying drawings.
As shown in FIG. 1, the intelligent control method for satellite power failure reconstruction based on the wild goat algorithm comprises the following steps:
the method comprises the following steps that S1, a hierarchical control system is designed, wherein the hierarchical control system comprises a mathematical physical layer and a network information layer which performs information interaction with the mathematical physical layer through a sensor, the mathematical physical layer comprises a sampling unit and a control unit, and the network information layer comprises a fault detection unit, a fault area detection unit, a fault reconstruction unit and an intelligent control agent unit; wherein:
the sampling unit and the control unit are used for carrying out information interaction with a network information layer through a sensor, and the information interaction comprises various electrical parameters of a satellite power supply system;
the fault detection is to perform information interaction with a sampling unit of a physical layer through a sensor, directly connect with a fault area detection and fault reconstruction unit and an intelligent control unit of a network layer, detect a current signal of a current circuit in real time and judge whether the current circuit works normally;
the fault area detection is connected with the fault detection and intelligent control unit, and when a circuit fault signal is received, the total load capacity of the fault area is judged and sent to the control unit; the circuit does not work when in normal operation;
the intelligent control unit is connected with the fault reconstruction, fault area detection and fault detection unit and has two working modes, when a circuit is in fault, a switch instruction is sent to each load line switch to realize the process of power supply fault reconstruction, then the reconstructed electrical quantity is sent to the sensor, the circuit is not reconstructed when being normal, and only information is transmitted with the sensor.
Sensor for measuring the shape of an objectThe physical quantity obtained by sampling the load output end by the processing layer is transmitted to the network information layer for data detection; the physical quantity obtained by sampling the load output end by the mathematical physical layer comprises the priority omega of N loads i Resistance R on the ith branch in different switch conducting states i Current I i Voltage V i Load L on ith bus i And the load state y i Load state y i The normal state is represented as 1 and the fault state is represented as 0.
S2, detecting sampling parameters, judging whether a fault occurs or not, if the signal abnormality is detected, judging that the fault occurs, positioning the fault, and respectively transmitting fault information to the S3 and the S4; and if the signal is normal, performing step S5, sending a normal signal to the intelligent control agent, and continuing working.
The process of detecting the sampling parameters comprises the following steps: according to the branch current I i Voltage V i And whether the values of the bus current and the bus voltage exceed the threshold values thereof, namely: I.C. A i ＞I imax And V i ＞V imax (ii) a When the detection unit detects that the electric quantity values of the system are normal, the detection unit judges that the line works in a normal state, and the detection unit sends a normal signal to the intelligent control unit; when the real-time detection signal is abnormal, the detection unit judges that the line has a fault, and respectively transmits an abnormal signal to the fault reconstruction unit in the step S3 to enable the fault reconstruction unit to execute a reconstruction function; failure information is transmitted to the failure region detection in step S4 to calculate the failure volume.
And S3, after receiving the fault signal and the fault positioning information, searching the optimal switch sequence of the system to reconstruct the fault based on the wild goat algorithm under the premise of considering the load grade, obtaining a reconstruction scheme of optimal power supply recovery under the load grade, and transmitting the reconstruction scheme to the intelligent control system in the step S5 to switch the switch.
The method comprises the following specific steps of searching the optimal switch sequence of the system for fault reconstruction on the premise of considering the load grade based on the wild goat algorithm:
step S31, composition of multi-objective function model
By usingThe linear weighting method considers the difference of each objective function value in dimension, order of magnitude and degree of importance through different weighting coefficients, basically solves the incomparable contradiction between feasible schemes, and according to the maximum recovery load f 1 And minimum power loss f 2 The importance degree of the fault recovery switch is used as weight, and a comprehensive objective function F is calculated, so that the multi-objective function can be effectively and accurately embodied, and an optimal fault recovery switch sequence is obtained.
The formula is as follows:
F＝λ 1 f 1 +λ 2 f 2
wherein, ω is i Is the priority, R, of the load i Is the resistance, L, of the ith branch in the on-state of different switches i Is the load on the ith bus, y i Is a load state, a load state y i The normal state is represented as 1, and the fault state is represented as 0, lambda 1 And λ 2 Is a constant;
step S32, constraint conditions of inequality: in the research of the optimization configuration problem of the satellite power supply fault, inequality constraints are considered besides equality constraint conditions, wherein the inequality constraint conditions comprise node voltage limit value constraints, line transmission power constraints and transmission line current constraints;
1) Node voltage limit constraints
After the satellite power supply is connected to the power distribution network, the local voltage of the power distribution network is out of limit due to the overlarge satellite power supply capacity, and therefore the voltage V of each node must be ensured i All relevant constraints are satisfied:
V imax ＞V i ＞V imin ；
2) Line transmission power constraint
The power supply setting constraint is that the transmission capacity of the transmission line is limited to a certain extent because the type of the power grid wire is fixed. Therefore, the transmission power of the line needs to be controlled within a certain range to prevent negative damage to the power system.
The transmission power of the line between the nodes meets the relevant constraint conditions as follows:
P ij ＜P ijmax ，
P ij representing the power, P, of the line transmission between nodes ijmax Is the maximum value of the transmission power of the line;
3) Transmission line current constraints
When the power supply system works abnormally, the current flowing in the line may exceed the maximum allowable load current, which causes the power system to malfunction: the current I flowing through the connected line is required to be influenced by the stable operation of the power system i And (6) carrying out constraint. Current I flowing in the line i The relevant constraints are satisfied as follows:
I imax ＞I i ＞I imin ；
step S33, initializing population: the load switches are numbered and binary coded, the switch states are represented by 0 and 1, the switch states are disconnected into 0 and closed into 1, and the switch states are combined to form a problem solution omega g i Wherein ω g i ＝f[x i1 ,x i2 ,...,x iN ]For example, the switch [ S ] of FIG. 2 1-1 ，S 1-2 ……S 1-n ]＝[0,0,...0]For each problem solution, perform an equal number N Gi Random grouping of (2);
step S34, evaluating the individuals: the weight of each independent solution is calculated according to the comprehensive objective function:
wherein N is var Is the dimension of each problem optimization, N wg Is the population membership, f (wg) i ) Is wg per individual i Weight value of F (wg) i ) Is the value of the integrated objective function;
then, grouping the individuals with equal numbers, and calculating the proportion of each individual in the group, wherein the weight of the optimal switching sequence is 1, and the others are 0-1, the individual with the high weight of the group is selected as a leader of the group, and the other individuals are followers;
step S35, group movement: after grouping, each group will move towards the best try direction, and the leader and follower in the group will have two different movement trends, as follows:
1) The leader will follow its own motion vector v i (t) and best try direction p i (t) carrying out group movement with followers in the collar group;
2) Direction v of movement of follower ik (t) there are three trends of influence: the follower's own motion vector and best try direction and the leader's motion direction;
the following is the calculation of the motion formulas for the leader and follower, respectively:
the leader: v i (t+1)＝ω×v i (t)+R×rand×(p i (t)-ωg i (t))，
Following the person: v ik (t+1)＝ω×v i (t)+R×rand×(p i (t)-ωg i (t))+W lk (t)×rand×(ωg lk (t)-ωg i (t))，
Where ω represents the inertial weight, R represents the learning coefficient, W lk Is the weight of the kth leader, and rand is a random parameter, typically taken as 0 or 1,v i (t) is the motion vector at time t, p i (t) is the best try direction,. Omega.g i (t) is the ith individual weight, [ omega ] g lk (t) is the overall composition of the population at time tAverage weight of the members, v ik (t) is the motion vector of the follower;
step S36, reevaluation: re-evaluating each group in each group after mobile updating, and re-calculating the weight of each independent individual and selecting a leader;
step S37, group cooperation: in the process of searching for the optimal switch sequence, updating the weights of individuals in the groups regularly inside the groups, comparing the leader weights among the groups, attracting the individual with the lowest weight in the lower group by the group with the higher weight and then reducing the weight of the group, wherein the weight of the group is calculated in the following way:
through the increase of the number of iterations, the group with low leader weight gradually loses its group members, even the leader, which is the best sequence for reconstruction when different groups composed of independent individuals are merged into the same group due to mutual attraction of the weights and only has one leader.
S4, when the fault signal is received, the fault area detection unit positions and calculates the total load capacity of the fault area detected by the fault information
And will load the total capacity
Transmitting to an intelligent control unit to provide line capacity constraint for the intelligent control system in the step S5;
step S5, after receiving the signals from the system, the intelligent control unit sends switching instructions to the lines of each load, and in a normal mode of the system: after receiving the normal signal of the step S2, the system does not adjust the switching sequence and keeps the current switching mode; in a system failure mode: after the fault reconstruction sequence of the step S3 is received, the operation of the step S4 is carried out to switch the optimal switching sequence sent by the line according to the load priority and the line capacity;
as shown in fig. 3, after receiving a system signal, the intelligent control unit may determine a failure mode, and if the system is normal, maintain the current switching mode, not change the system switching sequence, and continue to detect the current electrical quantity; if the system shows a fault, the following operations are performed:
after the optimal switching sequence after algorithm reconstruction is obtained in the fault mode, ARC is calculated i I.e. the available power capacity of the line shown in the switching sequence, which is then combined with the total capacity of the fault load
And comparing to judge whether the current line can support system recovery.
If it is satisfied with
Then the normal recovery of the system is carried out; if not, the priority load grade is high to recover, then a switch exchange table is established, the switch capacity of the nearby lines is calculated, the pressure of load recovery is transmitted to the nearby lines, and if the pressure of load recovery is supposed to be required to be opened on the K1 line in the optimal switch sequence, 1-m (m) lines are opened<N) switching, but because the line capacity does not meet the requirement, the loads with high priority levels are sorted in a descending order, the maximum capacity is recovered, the remaining unrepaired parts establish a switching table, the line power capacity of the K2 line is calculated, a possible recovery plan and load distribution are made, and the steps are sequentially carried out.
By establishing the switch table, recovery can be performed according to the load grade, a certain fault tolerance rate can be given to the reconstructed optimal switch sequence, and the situation that the optimal switch sequence falls into the local optimal predicament is prevented. And when the capacity of each line meets the requirement, but partial low-grade loads still have no power supply recovery, performing load shedding reconstruction.
And S6, after the fault reconstruction is completed, detecting a current electric quantity signal, transmitting the current electric quantity signal to a physical layer through a sensor to perform a new round of closed-loop control on the electric quantity of the line, and repeating the actions.
Claims (4)
1. The intelligent optimization control method for satellite power supply fault reconstruction based on the wild goat algorithm is characterized by comprising the following steps of: the method comprises the following steps:
the method comprises the following steps that S1, a hierarchical control system is designed, wherein the hierarchical control system comprises a mathematical physical layer and a network information layer which performs information interaction with the mathematical physical layer through a sensor, the mathematical physical layer comprises a sampling unit and a control unit, and the network information layer comprises a fault detection unit, a fault area detection unit, a fault reconstruction unit and an intelligent control agent unit; the sensor transmits physical quantity obtained by sampling the load output end by the mathematical physical layer to the network information layer for data detection;
s2, detecting sampling parameters, judging whether a fault occurs, if the signal is detected to be abnormal, judging that the fault occurs, positioning the fault, and respectively transmitting fault information to the S3 and the S4; if the signal is detected to be normal, the step S5 is carried out, a normal signal is sent to the intelligent control agent, and the intelligent control agent continues to work;
s3, after receiving the fault signal and the fault positioning information, searching the optimal switch sequence of the system to reconstruct the fault based on the wild goat algorithm under the premise of considering the load grade, obtaining a reconstruction scheme of optimal power restoration under the load grade, and transmitting the reconstruction scheme to the intelligent control system in the step S5 to switch the switch;
the method comprises the following specific steps of searching the optimal switch sequence of the system for fault reconstruction on the premise of considering the load grade based on the wild goat algorithm:
step S31, composition of multi-objective function model
Using linear weighting according to the maximum recovery load f 1 And minimum power loss f 2 The importance degree of the target function is used as weight, and a comprehensive objective function F is obtained through calculation, wherein the formula is as follows:
F＝λ 1 f 1 +λ 2 f 2
wherein, ω is i Is the priority, R, of the load i Is the resistance, L, on the ith branch in different switch on states i Is the load on the ith bus, y i Is a load state, a load state y i The normal state is represented as 1, and the fault state is represented as 0, lambda 1 And λ 2 Is a constant;
step S32, constraint conditions of inequality: the inequality constraint conditions comprise node voltage limit constraint, line transmission power constraint and transmission line current constraint;
1) Node voltage limit constraints
After the satellite power supply is connected to the power distribution network, the voltage V of each node i The relevant constraints are satisfied as follows:
V imax ＞V i ＞V imin ；
2) Line transmission power constraints
The transmission power of the line between the nodes meets the relevant constraint conditions as follows:
P ij ＜P ijmax ，
P ij representing the power, P, of the line transmission between nodes ijmax Is the maximum value of the line transmission power;
3) Transmission line current constraints
Current flowing in the lineI i The relevant constraints are satisfied as follows:
I imax ＞I i ＞I imin ；
step S33, initializing population: the load switches are numbered and binary coded, the switch states are represented by 0 and 1, the switch states are disconnected into 0 and closed into 1, and the switch states are combined to form a problem solution omega g i Wherein ω g i ＝f[x i1 ,x i2 ,...,x iN ]；
Step S34, evaluating the individuals: the weight of each independent solution is calculated according to the comprehensive objective function:
wherein N is var Is the dimension of each problem optimization, N wg Is the population membership, f (wg) i ) Is wg per individual i Weight value of F (wg) i ) Is the value of the integrated objective function;
then, performing equal-number grouping on the individuals, and then performing calculation on the proportion of each individual in the group, wherein the weight of the optimal switching sequence is 1, the other numbers are 0-1, the individual with the high weight of the group is selected as a leader of the group, and the other individuals are followers;
step S35, group movement: after grouping, each group will move towards the best try direction, and the leader and follower in the group will have two different movement trends, as follows:
1) The leader will follow its own motion vector v i (t) and best try direction p i (t) following in a collar groupThe person moves the group;
2) Direction v of movement of follower ik (t) there are three trends of influence: the follower's own motion vector and best try direction and the leader's direction of motion;
the following is the calculation of the motion formulas for the leader and follower, respectively:
the leader: v i (t+1)＝ω×v i (t)+R×rand×(p i (t)-ωg i (t))，
Following the person:
V ik (t+1)＝ω×v i (t)+R×rand×(p i (t)-ωg i (t))+W lk (t)×rand×(ωg lk (t)-ωg i (t))，
where ω represents the inertial weight, R represents the learning coefficient, W lk Is the weight of the kth leader, rand is a random parameter, taken as 0 or 1,v i (t) is the motion vector at time t, p i (t) is the best try direction,. Omega.g i (t) is the ith individual weight,. Omega.g lk (t) is the average weight of all members of the population at time t, v ik (t) is the motion vector of the follower;
step S36, reevaluation: re-evaluating each group in each group after mobile updating, and re-calculating the weight of each independent individual and selecting a leader;
step S37, group cooperation: in the process of searching for the optimal switching sequence, the weights of individuals in groups are updated regularly in the population, leader weights are compared among the populations, the group with high weight attracts the individual with the lowest weight in the lower group, and then the weight of the group is reduced, and the weight of the population is calculated as follows:
through the increase of the number of iterations, the group with low leader weight gradually loses its group members, even the leader, and when different groups consisting of independent individuals are combined into the same group due to the mutual attraction of the weights and only one leader is owned, the leader is the best sequence for reconstruction;
s4, when the fault signal is received, the fault area detection unit positions and calculates the total load capacity of the fault area detected by the fault information
And will load the total capacity
Transmitting to an intelligent control unit to provide line capacity constraint for the intelligent control system in the step S5;
step S5, after receiving the signals from the system, the intelligent control unit sends switching instructions to the lines of each load, and in a normal mode of the system: after receiving the normal signal of the step S2, the system does not adjust the switching sequence and keeps the current switching mode; in a system failure mode: after the fault reconstruction sequence of the step S3 is received, the operation of the step S4 is carried out to switch the optimal switching sequence sent by the line according to the load priority and the line capacity;
and S6, after the fault reconstruction is completed, detecting a current electric quantity signal, transmitting the current electric quantity signal to a physical layer through a sensor to perform a new round of closed-loop control on the electric quantity of the line, and repeating the actions.
2. The wild goat algorithm-based intelligent optimization control method for satellite power failure reconstruction as claimed in claim 1, wherein the intelligent optimization control method is characterized in that: in the step S1, the physical quantity obtained by sampling the load output end by the mathematical physical layer includes the priority ω of the N loads i Resistance R on the ith branch in different switch conducting states i Current I i Voltage V i Load L on ith bus i And load state y i Load state y i The normal state is represented as 1 and the fault state is represented as 0.
3. The wild goat algorithm-based intelligent optimization control method for satellite power failure reconstruction as claimed in claim 2, wherein: the process of detecting the sampling parameters in step S2 is as follows: according to the branch current I i Voltage V i And whether the values of the bus current and the bus voltage exceed the threshold values thereof, namely: i is i ＞I imax And V i ＞V imax (ii) a When the detection unit detects that the electric quantity values of the system are normal, the detection unit judges that the line works in a normal state, and the detection unit sends a normal signal to the intelligent control unit; when the real-time detection signal is abnormal, the detection unit judges that the line has a fault, and respectively transmits an abnormal signal to the fault reconstruction unit in the step S3 to enable the fault reconstruction unit to execute a reconstruction function; failure information is transmitted to the failure region detection in step S4 to calculate the failure capacity.
4. The wild goat algorithm-based intelligent optimization control method for satellite power failure reconstruction as claimed in claim 2, wherein: in the system failure mode in step S5, the intelligent control unit performs the following operations: calculating the available power capacity ARC of the circuit shown by the switching sequence after the optimal switching sequence of the system is obtained according to the step S3 i Then the ARC is applied i And total capacity of fault load
Comparing, and judging whether the current line can support system recovery;
if it is satisfied with
Then the normal recovery of the system is carried out; if the load is not satisfied, the load with high priority level is recovered, then a switch exchange table is established, the switch capacity of the nearby line is calculated, the pressure of load recovery is transmitted to the nearby line, and when the capacity of each line meets the requirement, but partial loads with low priority levels are not recovered by power supply, load shedding reconstruction is performed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210735650.5A CN115079571B (en) | 2022-06-27 | 2022-06-27 | Intelligent optimization control method for satellite power supply fault reconstruction based on wild goat algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210735650.5A CN115079571B (en) | 2022-06-27 | 2022-06-27 | Intelligent optimization control method for satellite power supply fault reconstruction based on wild goat algorithm |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115079571A CN115079571A (en) | 2022-09-20 |
| CN115079571B true CN115079571B (en) | 2022-12-30 |
Family
ID=83254741
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210735650.5A Active CN115079571B (en) | 2022-06-27 | 2022-06-27 | Intelligent optimization control method for satellite power supply fault reconstruction based on wild goat algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115079571B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106374456A (en) * | 2016-09-14 | 2017-02-01 | 国网天津市电力公司 | Power supply recovery method of power distribution network comprising distributed power supplies after fault |
| CN108711881A (en) * | 2018-05-08 | 2018-10-26 | 国网山东省电力公司青岛供电公司 | A kind of active distribution network fault recovery method containing distributed generation resource and micro-capacitance sensor |
| CN110994598A (en) * | 2019-11-26 | 2020-04-10 | 国家电网有限公司 | Multi-target power grid fault recovery method and device |
| CN113036823A (en) * | 2021-03-10 | 2021-06-25 | 海南电网有限责任公司电力科学研究院 | Distributed power distribution network optimization reconstruction method |
| CN113036760A (en) * | 2021-03-31 | 2021-06-25 | 国网上海市电力公司 | Method and system for recovering power grid load after disaster and computer readable medium |
| CN114336638A (en) * | 2022-01-28 | 2022-04-12 | 国家电网有限公司 | Medium-voltage active power distribution network dynamic reconstruction method based on bubbling model |
-
2022
- 2022-06-27 CN CN202210735650.5A patent/CN115079571B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106374456A (en) * | 2016-09-14 | 2017-02-01 | 国网天津市电力公司 | Power supply recovery method of power distribution network comprising distributed power supplies after fault |
| CN108711881A (en) * | 2018-05-08 | 2018-10-26 | 国网山东省电力公司青岛供电公司 | A kind of active distribution network fault recovery method containing distributed generation resource and micro-capacitance sensor |
| CN110994598A (en) * | 2019-11-26 | 2020-04-10 | 国家电网有限公司 | Multi-target power grid fault recovery method and device |
| CN113036823A (en) * | 2021-03-10 | 2021-06-25 | 海南电网有限责任公司电力科学研究院 | Distributed power distribution network optimization reconstruction method |
| CN113036760A (en) * | 2021-03-31 | 2021-06-25 | 国网上海市电力公司 | Method and system for recovering power grid load after disaster and computer readable medium |
| CN114336638A (en) * | 2022-01-28 | 2022-04-12 | 国家电网有限公司 | Medium-voltage active power distribution network dynamic reconstruction method based on bubbling model |
Non-Patent Citations (5)
| Title |
|---|
| Fault detection and reconstruction for micro-satellite power subsystem based on PCA;Bo Lee 等;《IEEE》;20101231;全文 * |
| Wild Goats Algorithm：An Evolutionary Algorithm to Solve the Real-World Optimization Problems;Alireza Shefaei 等;《IEEE》;20171231;全文 * |
| 主动配电网故障重构粒子群算法;矫德强 等;《长春工业大学学报》;20201231;第41卷(第6期);全文 * |
| 基于主从博弈理论的含分布式发电的主动配电网故障恢复方法;景皓 等;《智慧电力》;20211231;第49卷(第5期);全文 * |
| 基于改进蚁群算法的配电网故障后恢复重构;卢志刚 等;《人工智能在电力系统中的应用》;20081231;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115079571A (en) | 2022-09-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110086168B (en) | Power distribution network operation topology identification method based on multi-source data fusion | |
| WO2022016622A1 (en) | Adaptive optimization and control method in event of failure of true bipolar flexible direct-current power transmission system | |
| CN109378815B (en) | Power distribution network closed loop switching power supply control method, device and equipment | |
| CN110311398B (en) | Connection topology, control system and method of novel energy storage battery system | |
| CN101763034B (en) | Method for realizing self-adapting network reconfiguration at distribution network terminal | |
| CN111697566A (en) | Reliability assessment method for active power distribution network information physical system considering information failure | |
| CN108931709A (en) | A kind of distribution network fault location method of high serious forgiveness | |
| CN111416359A (en) | Power distribution network reconstruction method considering weighted power flow entropy | |
| CN110350511B (en) | Power distribution network fault recovery method based on transfer power affinity | |
| CN109034467A (en) | A kind of electric system disaster-resistant type bulk transmission grid Multipurpose Optimal Method | |
| CN106711966A (en) | Distribution network protection fixed value setting method and system | |
| Wang et al. | A minimum hitting set algorithm with prejudging mechanism for model-based fault diagnosis in distribution networks | |
| Guan et al. | A multi-agent-based self-healing framework considering fault tolerance and automatic restoration for distribution networks | |
| CN111682525A (en) | Load transfer method based on optimal flow method and Mayeda spanning tree method | |
| Wang et al. | Multi-agent and ant colony optimization for ship integrated power system network reconfiguration | |
| CN115079571B (en) | Intelligent optimization control method for satellite power supply fault reconstruction based on wild goat algorithm | |
| CN113922358A (en) | Method for rapidly evaluating reliability of time sequence operation of power distribution network containing distributed energy storage | |
| CN108494088A (en) | Consider the intelligent distribution type feeder automation terminal of ad hoc network equipment state | |
| CN117791560A (en) | Active power distribution network elastic self-healing method considering dynamic micro-grid and controller | |
| Patel et al. | A hopfield neural network based reconfiguration algorithm for power distribution systems | |
| CN116993324A (en) | Fault recovery optimization method and system for micro-active power distribution network | |
| CN115408866A (en) | Time-varying topological structure-based constellation on-orbit health assessment method and terminal | |
| CN108090616A (en) | A kind of electric system Active Splitting optimal section searching method | |
| CN113740650A (en) | Ship power system fault detection method, terminal device and storage medium | |
| Nguyen et al. | Using an improved Neural Network with Bacterial Foraging Optimization algorithm for Load Shedding |
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 |