CN114253736A - Novel intelligent distributed computing and operating system for power grid - Google Patents
Novel intelligent distributed computing and operating system for power grid Download PDFInfo
- Publication number
- CN114253736A CN114253736A CN202210121383.2A CN202210121383A CN114253736A CN 114253736 A CN114253736 A CN 114253736A CN 202210121383 A CN202210121383 A CN 202210121383A CN 114253736 A CN114253736 A CN 114253736A
- Authority
- CN
- China
- Prior art keywords
- data
- calculation
- server
- agent
- scheduling server
- 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.)
- Withdrawn
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5083—Techniques for rebalancing the load in a distributed system
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention discloses an intelligent distributed computation and operation system for a power grid, which comprises: the system comprises an initial state configuration module, an online starting module, a data period calculation module, an agent management module and a distributed operation processing module. According to the intelligent distributed computation and operation system for the power grid, the data gateway, the scheduling server, the data server and the computing nodes are matched to complete the starting computation and the agent of the distributed computation of the power grid, the operation management of the computing nodes based on simulation and analysis is carried out, and the overall architecture and the implementation are in a loosely-coupled and loosely-associated state. The parallel distributed large-scale high-performance computing requirement of the power system enterprise level is met.
Description
Technical Field
The invention relates to the field of distributed computing, in particular to an intelligent distributed computing and operating system for a power grid.
Background
The large-scale distributed parallel computing general platform makes full use of idle computing capacity of cluster nodes in the power system or worker computers in the power mechanism, tasks submitted by users in batches are respectively dispatched to the participating node computers for execution, and computing input files and result files are subjected to data exchange through a local area network or a wide area network.
The distributed computing projects which are common abroad are mainly applied to analyzing the radio signals outside the land, so that the life signs outside the land, gene sequencing, the internal structure of the protein, and the like are searched, analyzed and computed. These projects are typically designed to address the serious lack of computing power, but there is no enterprise-level requirement for flexibility in computing tasks, computing power, and parallel computing efficiency.
Disclosure of Invention
In order to solve the defects and shortcomings in the prior art, the invention provides an intelligent distributed computing and operating system for a power grid, which comprises: the system comprises an initial state configuration module, an online starting module, a data period calculation module, an agent management module and a distributed operation processing module; wherein.
And the initial state configuration module is used for configuring a data gateway, a scheduling server, a data server and the computer node cluster, entering a ready state after the data gateway, the scheduling server, the data server and the computer node cluster are started, and waiting for a human-computer interface to send a starting instruction.
The online starting module sends a starting command to the main scheduling server through the human-computer interface, and the main scheduling server receives the starting command and returns a starting success message to the human-computer interface.
The data period calculation module is called by the scheduling server to perform multi-period calculation management, the scheduling server rebroadcasts the load flow data to the calculation nodes to perform calculation, then receives calculation results returned by the calculation nodes, and performs analysis according to the collected calculation results to generate a new periodic calculation task; the new calculation task is dynamically distributed to the calculation nodes for secondary calculation, and the secondary calculation result is returned, analyzed and regenerated into a new task; this phase is repeated for a number of times until no new computing tasks are generated, so that a data cycle is completed.
The phase calculation module is used for performing phase calculation in the data period calculation, and in the data period calculation process, the scheduling server calls the phase calculation module to perform analysis and processing after receiving a calculation result uploaded by the calculation node, and generates a new phase calculation task. The new stage computing task information includes task input or setup files, task computing node IP addresses, and the like. The scheduling server inputs the task to the computing node according to the task node IP, or sets a file to be unicast or multicast to the computing node through a logic group, and the computing node can trigger new computation after receiving the task or the file. And after the calculation is finished, unicasting the phase calculation result to a scheduling server. And then the scheduling server calls a phase calculation module to generate a staged calculation task of the next phase according to the received phase calculation result, and the new staged calculation task is dynamically distributed to the calculation nodes to carry out the staged calculation.
The agent management module is used for executing and coordinating and controlling distributed computing and comprises.
And the tasks of the execution-level agents comprise monitoring whether the voltage of the computing node exceeds the limit, sending a control request to the coordination-level agents, receiving a control instruction issued by the coordination-level agents and implementing so as to realize voltage detection and control. And the coordination level agents receive the control requests and the operation data sent by the execution level agents, communicate with other coordination level agents and exchange data, calculate the optimal control quantity in each region and then send the optimal control quantity to the execution level agents so as to realize coordination calculation and decision.
The distributed operation processing module is used for: and performing transient stability simulation based on the fault. And carrying out power supply operation analysis of the power grid network. And performing operation implementation of the fault model prediction control.
Preferably, in the online starting process, the master scheduling server starts to forward and process the formatted power flow data file received from the data gateway after the start. If the main scheduling server breaks down while the human-computer interface sends the instruction, the human-computer interface sends the instruction overtime and makes an error, and the human-computer interface retries the operation until the operation is successfully returned after obtaining the error prompt information.
Preferably, in the data period calculation process, if the first data period has not ended yet, that is, the calculation has not been completed at this time, and a situation that the next period has started occurs, that is, the scheduling server receives new power flow data again at this time, the data period calculation module controls the calculation to directly enter the next period, and the data and state of the previous period are discarded to enter the new data period calculation. And the moment when the data gateway issues the format trend data is taken as the beginning of a data period.
Preferably, the load flow data is collected by the data gateway and is periodically sent to the scheduling server through the multicast channel in the form of a format file, the scheduling server is in a ready state after being started, the load flow data is sent to the computing nodes in the cluster after a trigger signal of the human-computer interface is obtained, and the computing nodes feed back the computed result to the scheduling server. The scheduling server transmits the calculation result to the data server through the multicast message channel. After the data server obtains the task data of the data gateway, a subdirectory is established locally, and the task data is stored on a hard disk in a file format; meanwhile, the calculation result data are classified according to periods, and the calculation result data and the task data are stored in the same subdirectory. A backup of the data server also exists, and the backup data server functions as: the data and state synchronization with the main data server is always maintained, and when the main data server fails, the backup data server takes over the work of the main server.
Preferably, the scheduling server is provided with a backup system, i.e. a backup scheduling server. If the main dispatching server fails in operation, the work of the main dispatching server is taken over by the backup dispatching server. The method of succession is: tasks that were not completed in the previous data cycle are no longer resumed and the resulting intermediate data and state are discarded, taking over the scheduled tasks from the beginning of the next data cycle. The master-slave fault switching time is less than 30 s.
Preferably, within each coordination level agent, two subagents are included, one being a sending agent responsible for sending data and the other being a receiving agent responsible for receiving data. The sending agent is automatically generated after each calculation is finished, and the receiving host with the known IP address is obtained after the sending agent is generated. The sending agent searches for a receiving agent in the receiving host. The sending agent takes the calculation result converted into the character string as the content of the information, and the address of the receiving agent as the receiving address of the information. The sending agent and the receiving agent are started before transient stability simulation, the sending agent and the receiving agent run through the whole life cycle of the coordination level agent after being started, the receiving agent monitors whether new information arrives all the time, and if the information is received, the content of the information is returned as a return value and continues to be monitored.
Preferably, the fault-based transient stability simulation includes.
(1) And configuring the maximum iteration times, and calculating the transient stability margin of the power grid in each fault category under the reference tide before network adjustment.
(2) If multiple fault types cause the transient power of the power grid to be unstable, setting a limit power adjusting mode in a given candidate measure space by using the participation factors of the stabilizing unit for multiple faults of the same type, executing limit calculation, and outputting a limit power calculation result.
(3) And for a plurality of faults of each kind, carrying out optimal solution on the control measure by taking the minimum control cost as a target in a given candidate measure space based on the limit power calculation result by utilizing the transient safety constraint.
(4) Applying the control measures obtained by the optimization solution to the reference tidal current data to obtain the adjusted tidal current data, performing transient security check on the adjusted tidal current data, and if the transient is stable under each fault type, the control measures are effective; and (4) judging whether the maximum iteration number is reached, if so, exiting, and otherwise, turning to the step (1).
Wherein the input data of the transient stability simulation comprises: input data for online safety and stability evaluation; generator, load, reactive compensation candidate measure space; and a tie line or tie profile definition.
The output data of the transient stability simulation includes: preventive stability control measures; the adjusted power grid operation mode data; initial power and limit power of the tie line or tie section; conflict fault control information; and overall control costs.
According to the intelligent distributed computation and operation system for the power grid, the data gateway, the scheduling server, the data server and the computing nodes are matched to complete the starting computation and the agent of the distributed computation of the power grid, the operation management of the computing nodes based on simulation and analysis is carried out, and the overall architecture and the implementation are in a loosely-coupled and loosely-associated state. The parallel distributed large-scale high-performance computing requirement of the power system enterprise level is met.
Drawings
Fig. 1 is a structural diagram of an intelligent distributed computing and operating system of a power grid according to the present invention.
Detailed Description
As shown in fig. 1, the present invention provides an intelligent distributed computing and operating system for a power grid, including: the system comprises an initial state configuration module, an online starting module, a data period calculation module, an agent management module and a distributed operation processing module; wherein.
And the initial state configuration module is used for configuring a data gateway, a scheduling server, a data server and the computer node cluster, entering a ready state after the data gateway, the scheduling server, the data server and the computer node cluster are started, and waiting for a human-computer interface to send a starting instruction. And in the process of entering the ready state, the data gateway issues the formatted tidal current data file periodically, and the main and standby scheduling servers directly discard the formatted tidal current data file after receiving the formatted tidal current data file and do not forward the formatted tidal current data file and do not process the formatted tidal current data file.
The online starting module sends a starting command to the main scheduling server through the human-computer interface, and the main scheduling server receives the starting command and returns a starting success message to the human-computer interface. And the main scheduling server starts to forward and process the format power flow data file received from the data gateway after starting. If the main scheduling server breaks down while the human-computer interface sends the instruction, the human-computer interface sends the instruction overtime and makes an error, and the human-computer interface retries the operation until the operation is successfully returned after obtaining the error prompt information.
The data period calculation module is called by the scheduling server to perform multi-period calculation management, the scheduling server rebroadcasts the load flow data to the calculation nodes to perform calculation, then receives calculation results returned by the calculation nodes, and performs analysis according to the collected calculation results to generate a new periodic calculation task; the new calculation task is dynamically distributed to the calculation nodes for secondary calculation, and the secondary calculation result is returned, analyzed and regenerated into a new task; this phase is repeated for a number of times until no new computing tasks are generated, so that a data cycle is completed. If the first data period is not finished, namely the calculation is not finished at the moment, and the situation that the next period is started occurs, namely the scheduling server receives new load flow data again at the moment, the data period calculation module controls the calculation to directly enter the next period, the data and the state of the previous period are discarded, and the calculation enters the new data period. And the moment when the data gateway issues the format trend data is taken as the beginning of a data period.
The load flow data is collected by the data gateway and is periodically sent to the scheduling server through the multicast channel in the form of a format file, the scheduling server is in a ready state after being started, the load flow data is sent to the computing nodes in the cluster after a trigger signal of the man-machine interface is obtained, and the computing nodes feed back the computed result to the scheduling server. The scheduling server transmits the calculation result to the data server through the multicast message channel. After the data server obtains the task data of the data gateway, a subdirectory is established locally, and the task data is stored on a hard disk in a file format; meanwhile, the calculation result data are classified according to periods, and the calculation result data and the task data are stored in the same subdirectory. A backup of the data server also exists, and the backup data server functions as: the data and state synchronization with the main data server is always maintained, and when the main data server fails, the backup data server takes over the work of the main server.
Meanwhile, a backup system, namely a backup scheduling server, is provided for the scheduling server. If the main dispatching server fails in operation, the work of the main dispatching server is taken over by the backup dispatching server. The method of succession is: tasks that were not completed in the previous data cycle are no longer resumed and the resulting intermediate data and state are discarded, taking over the scheduled tasks from the beginning of the next data cycle. The master-slave fault switching time is less than 30 s.
The phase calculation module is used for performing phase calculation in the data period calculation, and in the data period calculation process, the scheduling server calls the phase calculation module to perform analysis and processing after receiving a calculation result uploaded by the calculation node, and generates a new phase calculation task. The new stage computing task information includes task input or setup files, task computing node IP addresses, and the like. The scheduling server inputs the task to the computing node according to the task node IP, or sets a file to be unicast or multicast to the computing node through a logic group, and the computing node can trigger new computation after receiving the task or the file. And after the calculation is finished, unicasting the phase calculation result to a scheduling server. And then the scheduling server calls a phase calculation module to generate a staged calculation task of the next phase according to the received phase calculation result, and the new staged calculation task is dynamically distributed to the calculation nodes to carry out the staged calculation. The complete steps of the phase calculation include: generating, distributing, calculating, collecting results and generating next-stage tasks. The steps are repeated and superimposed in stages until no new computing tasks are generated.
The agent management module is used for executing and coordinating and controlling distributed computing and comprises.
And the tasks of the execution-level agents comprise monitoring whether the voltage of the computing node exceeds the limit, sending a control request to the coordination-level agents, receiving a control instruction issued by the coordination-level agents and implementing so as to realize voltage detection and control. And the coordination level agents receive the control requests and the operation data sent by the execution level agents, communicate with other coordination level agents and exchange data, calculate the optimal control quantity in each region and then send the optimal control quantity to the execution level agents so as to realize coordination calculation and decision.
At the level of coordination level proxy, because the sizes of the regions are different from the configuration of the computer, the calculation time is different, and the regions cannot be accurately calculated due to asynchronous calculation. Therefore, the above problem is handled by adopting the principle of "fast and slow", namely, the region with fast computation needs to suspend the computation thread until receiving a new iteration result sent by other agents.
Each coordination level agent comprises two sub-agents, wherein one sub-agent is a sending agent responsible for sending data, and the other sub-agent is a receiving agent responsible for receiving data. The sending agent is automatically generated after each calculation is finished, and the receiving host with the known IP address is obtained after the sending agent is generated. The sending agent searches for a receiving agent in the receiving host. The sending agent takes the calculation result converted into the character string as the content of the information, and the address of the receiving agent as the receiving address of the information. The sending agent and the receiving agent are started before transient stability simulation, the sending agent and the receiving agent run through the whole life cycle of the coordination level agent after being started, the receiving agent monitors whether new information arrives all the time, and if the information is received, the content of the information is returned as a return value and continues to be monitored.
The distributed operation processing module is used for processing the operation data.
And performing transient stability simulation based on the fault.
(1) And configuring the maximum iteration times, and calculating the transient stability margin of the power grid in each fault category under the reference tide before network adjustment. The fault types comprise short circuit faults, short line faults, overload, line overload, low voltage, frequency oscillation and the like.
(2) If multiple fault types cause the transient power of the power grid to be unstable, setting a limit power adjusting mode in a given candidate measure space by using the participation factors of the stabilizing unit for multiple faults of the same type, executing limit calculation, and outputting a limit power calculation result.
(3) And for a plurality of faults of each kind, carrying out optimal solution on the control measure by taking the minimum control cost as a target in a given candidate measure space based on the limit power calculation result by utilizing the transient safety constraint.
(4) Applying the control measures obtained by the optimization solution to the reference tidal current data to obtain the adjusted tidal current data, performing transient security check on the adjusted tidal current data, and if the transient is stable under each fault type, the control measures are effective; and (4) judging whether the maximum iteration number is reached, if so, exiting, and otherwise, turning to the step (1).
Wherein the input data of the transient stability simulation comprises: input data for online safety and stability evaluation; generator, load, reactive compensation candidate measure space; and a tie line or tie profile definition.
The output data of the transient stability simulation includes: preventive stability control measures; the adjusted power grid operation mode data; initial power and limit power of the tie line or tie section; conflict fault control information; and overall control costs.
And carrying out power supply operation analysis of the power grid network.
1) And configuring a power grid network structure. The power grid network structure comprises a power grid full-wiring mode, and a peak operation mode and a maximum peak operation mode are adopted. And configuring the unit and the load data. The unit and load data comprise power generation data of a current power grid unit and a standby power grid unit, peak load and maximum peak load data of the current power grid, and peak load and maximum peak load data of the standby power grid.
2) And carrying out distributed stable calculation analysis. And performing power grid safety and stability analysis aiming at the network, the generator data and the load data of the power grid, wherein the safety and stability analysis comprises stable calculation in three aspects, namely power angle stability, thermal stability and voltage stability. For power angle stability and thermal stability, performing stability calculation analysis by a power grid operator by using a power grid control section, wherein the control section comprises thermal stability and transient stability, and a larger limit value in the thermal stability and the transient stability is used as a power flow limit of the stability control section so that a power grid dispatching operator controls power flow distribution of a power grid; for voltage stabilization, which is associated with reactive power, this is done by installing reactive power supplies nearby, including in-situ installation of shunt capacitors, static compensators SVC, etc.
At present, no matter from the viewpoint of power grid planning or from the viewpoint of actual dispatching operation of a power grid, voltage stabilization is not considered as a decisive factor in a power grid network structure. Transient stability and thermal stability are the most important considerations in the planning of the power grid, and therefore the impact of voltage stability on the structure and planning of the power grid is not elaborated.
3) And for the power transmission network in the looped network form, judging whether the thermal stability limit or the transient stability limit is larger than the transmission power. The stably calculated limit includes both the limit of power transmission to the power receiving area and the transmission power of the transmission section between the ring networks or between the areas. If the thermal stability limit or transient stability limit is not larger than the transmission power, the output stability limit does not meet the requirement of the transmission power, and then the 1) network is adjusted, otherwise, the 4) network is entered.
4) Judging whether the power grid outputs a receiving-end power grid and a thermal stability limit, if so, turning to 5) judging whether the power grid has a structural problem; and if the output is transient stability limit or non-receiving end power grid, the power grid meets the requirement that the limit is larger than the transmission power, and the analysis is finished.
5) When the receiving-end power grid and the thermal stability limit are output, judging whether the transmission power limit is larger than the power load, and if the transmission limit is larger than the power load, ending the analysis; otherwise, configuration control is carried out on a receiving end through a power grid control measure, the power grid control measure is determined by transient stability simulation based on faults, or a default control measure can be directly adopted, and the default control measure comprises load shedding and a network structure for configuring a low-level power grid. And if the transmission power limit is still smaller than the power load of the receiving-end power grid after the receiving-end power grid is configured and controlled by a power grid adjusting means, performing network optimization on the power grid to readjust the network, and then turning to 1).
The optimization of the power grid network is the prior art and is not described herein again.
The fault model predictive control operation is implemented by the following specific steps.
(1, judging whether the simulation power grid branch meets opportunity constraint conditions, and entering (2) when the simulation power grid branch meets the opportunity constraint conditions, or exiting.
(2 at time T0 after the fault, voltage was detected at a lower level, the MPC model predictive control performed the action, let Tk = T0, recording the first sampling time.
(3, performing time domain simulation in (Tk, Tk + 1), and calculating the track sensitivity at the Tk +1 moment.
(4, judging whether the sampling interval exceeds the control time domain, if not, solving the simplified quadratic programming model, obtaining the estimated control quantity of the Tk moment, applying the estimated control quantity to the system, enabling k = k +1, entering the next sampling interval, and returning to the step (3), if so, calculating the track sensitivity of the rest sampling moments in the prediction time domain.
And (5) returning Tk = T0 to the first sampling time of the whole prediction time domain, and performing time domain simulation in the prediction time domain (Tk, Tk + NTs) to obtain a system track of the power grid.
Let k = k + l, T0= Tk, i.e. the whole prediction horizon moves forward by one sampling interval, return to step (2, perform the next rolling optimization until all sampling intervals are traversed.
And judging whether the simulated power grid branch meets opportunity constraint conditions or not, including.
A1. Given the allowed number of random simulations Nmax.
A2. The chance constraint holds true counter N = 0.
A3. And acquiring all the calculation nodes i with the newly added power supply, acquiring one calculation node i, and calculating the branch power P of the calculation node i.
A4. When P < = Pmax, set N = N + 1. And Pmax is the branch power transmission capacity. Otherwise, the next computing node i is taken, and the step A3 is returned.
A5. If N/Nmax > = a, a is the confidence level of the line power constraint, the simulation meets the opportunity constraint, and the simulation is ended; and otherwise, taking the next computing node i, returning to the step A3, if all the computing nodes i are acquired at the moment, the simulation does not meet the opportunity constraint, and exiting the simulation.
The initial state configuration module, the online starting module, the data period calculation module, the agent management module and the distributed operation processing module realize module functions through corresponding processes and threads, the modules are connected with each other through logic or electricity, and the specific functions of the modules can be realized by program codes stored in a physical processor execution memory.
According to the intelligent distributed computation and operation system for the power grid, the data gateway, the scheduling server, the data server and the computing nodes are matched to complete the starting computation and the agent of the distributed computation of the power grid, the operation management of the computing nodes based on simulation and analysis is carried out, and the overall architecture and the implementation are in a loosely-coupled and loosely-associated state. The parallel distributed large-scale high-performance computing requirement of the power system enterprise level is met.
The above embodiments are only exemplary embodiments of the present invention, and are not intended to limit the present invention, and the scope of the present invention is defined by the claims. Various modifications and equivalents may be made by those skilled in the art within the spirit and scope of the present invention, and such modifications and equivalents should also be considered as falling within the scope of the present invention.
Claims (7)
1. The utility model provides a power grid intelligent distributed computation and operating system which characterized in that includes: the system comprises an initial state configuration module, an online starting module, a data period calculation module, an agent management module and a distributed operation processing module; wherein the content of the first and second substances,
the initial state configuration module configures a data gateway, a scheduling server, a data server and the computer node cluster to enter a ready state after being started, and waits for a human-computer interface to send a starting instruction;
the online starting module sends a starting command to the main scheduling server through the human-computer interface, and the main scheduling server receives the starting command and returns a starting success message to the human-computer interface;
the data period calculation module is called by the scheduling server to perform multi-period calculation management, the scheduling server rebroadcasts the load flow data to the calculation nodes to perform calculation, then receives calculation results returned by the calculation nodes, and performs analysis according to the collected calculation results to generate a new periodic calculation task; the new calculation task is dynamically distributed to the calculation nodes for secondary calculation, and the secondary calculation result is returned, analyzed and regenerated into a new task; repeating the steps for multiple times until no new calculation task is generated, and ending one data period;
the phase calculation module is used for performing phase calculation in data period calculation, and in the data period calculation process, the scheduling server calls the phase calculation module to perform analysis and processing after receiving a calculation result uploaded by a calculation node, and generates a new phase calculation task; the new stage calculation task information comprises a task input or setting file, a task calculation node IP address and the like; the scheduling server inputs tasks to the computing nodes according to the task node IP, or sets files to be unicast or multicast to the computing nodes through a logic group, and the computing nodes can trigger new computation after receiving the tasks or the files; after the calculation is finished, unicasting a stage calculation result to a scheduling server; then, the scheduling server calls a stage calculation module to generate a stage calculation task of the next stage according to the received stage calculation result, and the new stage calculation task is dynamically distributed to the calculation nodes to carry out stage calculation;
the agent management module is used for executing and coordinating and controlling distributed computing, and comprises:
the task of the execution-level agent comprises monitoring whether the voltage of the computing node is out of limit or not, sending a control request to the coordination-level agent, receiving a control instruction issued by the coordination-level agent and implementing the control instruction so as to realize voltage detection and control; the coordination level agents are used for receiving the control requests and the operation data sent by the execution level agents, communicating with other coordination level agents, exchanging data, calculating the optimal control quantity in each area and then sending the optimal control quantity to the execution level agents to realize coordination calculation and decision;
the distributed operation processing module is used for: performing transient stability simulation based on the fault; carrying out power supply operation analysis of a power grid network; and performing operation implementation of the fault model prediction control.
2. The system of claim 1, wherein during the on-line startup process, the main scheduling server starts to forward and process the formatted power flow data file received from the data gateway after the startup; if the main scheduling server breaks down while the human-computer interface sends the instruction, the human-computer interface sends the instruction overtime and makes an error, and the human-computer interface retries the operation until the operation is successfully returned after obtaining the error prompt information.
3. The system according to claim 1, wherein in the data period calculation process, if the first data period has not been finished, that is, the calculation has not been finished, and a situation that the next period has already started occurs, that is, the scheduling server receives new load flow data again, the data period calculation module controls the calculation to directly enter the next period, and the data and state of the previous period are discarded and the calculation enters the new data period; and the moment when the data gateway issues the format trend data is taken as the beginning of a data period.
4. The system of claim 3, wherein the load flow data is collected by the data gateway and is periodically transmitted to the scheduling server through a multicast channel in the form of a format file, the scheduling server is in a ready state after being started, the scheduling server starts to transmit the load flow data to the computing nodes in the cluster after obtaining a trigger signal of the human-computer interface, and the computing nodes feed back the computed result to the scheduling server; the scheduling server transmits the calculation result to the data server through the multicast message channel; after the data server obtains the task data of the data gateway, a subdirectory is established locally, and the task data is stored on a hard disk in a file format; meanwhile, the calculation result data are classified according to periods, and are stored in the same subdirectory with the task data; a backup of the data server also exists, and the backup data server functions as: the data and state synchronization with the main data server is always maintained, and when the main data server fails, the backup data server takes over the work of the main server.
5. A system according to claim 3, characterized in that the scheduling server is provided with a backup system, i.e. a backup scheduling server; if the main dispatching server fails in operation, the work of the main dispatching server is taken over by the backup dispatching server; the method of succession is: tasks which are not completed in the previous data period are not continued any more, generated intermediate data and states are discarded, and scheduled tasks are replaced from the beginning of the next data period; the master-slave fault switching time is less than 30 s.
6. The system of claim 1, wherein each coordination level agent comprises two sub-agents, one of which is a sending agent responsible for sending data and the other of which is a receiving agent responsible for receiving data; the sending agent is automatically generated after each calculation is finished, and a receiving host with a known IP address is obtained after the sending agent is generated; the sending agent searches for a receiving agent in the receiving host; the sending agent takes the calculation result converted into the character string as the content of the information, and the address of the receiving agent is taken as the receiving address of the information; the sending agent and the receiving agent are started before transient stability simulation, the sending agent and the receiving agent run through the whole life cycle of the coordination level agent after being started, the receiving agent monitors whether new information arrives all the time, and if the information is received, the content of the information is returned as a return value and continues to be monitored.
7. The system of claim 1, wherein the fault-based transient stability simulation comprises:
(1) configuring the maximum iteration times, and calculating the transient stability margin of the power grid in each fault category under the reference tide before network adjustment;
(2) if multiple fault types cause the transient power of the power grid to be unstable, setting a limit power adjustment mode in a given candidate measure space by using the participation factors of the stabilizing unit for multiple faults of the same type, executing limit calculation, and outputting a limit power calculation result;
(3) for a plurality of faults of each kind, optimizing and solving the control measure by using transient safety constraint and taking the minimum control cost as a target in a given candidate measure space based on a limit power calculation result;
(4) applying the control measures obtained by the optimization solution to the reference tidal current data to obtain the adjusted tidal current data, performing transient security check on the adjusted tidal current data, and if the transient is stable under each fault type, the control measures are effective; judging whether the maximum iteration times is reached, if so, exiting, otherwise, turning to the step (1);
wherein the input data of the transient stability simulation comprises: input data for online safety and stability evaluation; generator, load, reactive compensation candidate measure space; and a connection line or a connection section definition;
the output data of the transient stability simulation includes: preventive stability control measures; the adjusted power grid operation mode data; initial power and limit power of the tie line or tie section; conflict fault control information; and overall control costs.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110853701.XA CN113419866A (en) | 2021-07-28 | 2021-07-28 | Intelligent distributed computation and operation system for power grid |
| CN202110853701X | 2021-07-28 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114253736A true CN114253736A (en) | 2022-03-29 |
Family
ID=77718442
Family Applications (2)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110853701.XA Withdrawn CN113419866A (en) | 2021-07-28 | 2021-07-28 | Intelligent distributed computation and operation system for power grid |
| CN202210121383.2A Withdrawn CN114253736A (en) | 2021-07-28 | 2022-02-09 | Novel intelligent distributed computing and operating system for power grid |
Family Applications Before (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110853701.XA Withdrawn CN113419866A (en) | 2021-07-28 | 2021-07-28 | Intelligent distributed computation and operation system for power grid |
Country Status (1)
| Country | Link |
|---|---|
| CN (2) | CN113419866A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115981610A (en) * | 2023-03-17 | 2023-04-18 | 科大国创软件股份有限公司 | Comprehensive operation platform of photovoltaic energy storage system based on Lua script |
-
2021
- 2021-07-28 CN CN202110853701.XA patent/CN113419866A/en not_active Withdrawn
-
2022
- 2022-02-09 CN CN202210121383.2A patent/CN114253736A/en not_active Withdrawn
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115981610A (en) * | 2023-03-17 | 2023-04-18 | 科大国创软件股份有限公司 | Comprehensive operation platform of photovoltaic energy storage system based on Lua script |
| CN115981610B (en) * | 2023-03-17 | 2023-06-02 | 科大国创软件股份有限公司 | Comprehensive operation platform of photovoltaic energy storage system based on Lua script |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113419866A (en) | 2021-09-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107330056B (en) | Wind power plant SCADA system based on big data cloud computing platform and operation method thereof | |
| CN102063336B (en) | Distributed computing multiple application function asynchronous concurrent scheduling method | |
| CN108845878A (en) | The big data processing method and processing device calculated based on serverless backup | |
| CN112053041B (en) | Province-region integrated source network load storage real-time control system and control method | |
| CN103823541B (en) | Virtual data center energy-saving scheduling method and equipment | |
| CN101685480A (en) | Parallel computing method for security and stability analysis of large power grid and computing platform | |
| CN103297505A (en) | Multi-energy online control method and system for data center under dynamic cloud service request | |
| CN105024939A (en) | OpenFlow-based distributed controller system in SDN network environment | |
| CN114253736A (en) | Novel intelligent distributed computing and operating system for power grid | |
| US20170244252A1 (en) | Autonomous Operational Platform for Micro-Grid Energy Management | |
| Toersche et al. | Energy management with TRIANA on FPAI | |
| CN106936905B (en) | Openstack-based scheduling method and scheduling system of Nova component virtual machine | |
| US20220198584A1 (en) | Computer-Assisted Energy Management Method And Energy Management System | |
| CN114265683A (en) | New system for power task allocation, calculation and management | |
| CN104298536A (en) | Dynamic frequency modulation and pressure adjustment technology based data center energy-saving dispatching method | |
| CN111404195B (en) | Intelligent gateway-based scheduling method for microgrid with distributed power supply | |
| CN104580498A (en) | Self-adaptive cloud management platform | |
| KR102334975B1 (en) | System For Controlling An Automate Equipment | |
| Paya et al. | Cloud-based Simulation of a Smart Power Grid | |
| CN110490763B (en) | Peak regulation assisting method and system | |
| KR20180137622A (en) | Apparatus for remote managing new renewable energy plant | |
| Sun et al. | Probability-Based Online Algorithm for Switch Operation of Energy Efficient Data Center | |
| Peng et al. | Digital Simulation and Automatic Test of Source Network Load Storage Cooperative Control | |
| Maaouia et al. | Towards optimizing energy consumption in cloud | |
| Li et al. | Node Resource Balance Scheduling Algorithm of Power Internet of Things Based on Big Data |
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 | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20220329 |
|
| WW01 | Invention patent application withdrawn after publication |