CN102681442A - Intelligent fault-tolerant control system and control method for zonal power distribution of full-electric boat - Google Patents
Intelligent fault-tolerant control system and control method for zonal power distribution of full-electric boat Download PDFInfo
- Publication number
- CN102681442A CN102681442A CN2012101702375A CN201210170237A CN102681442A CN 102681442 A CN102681442 A CN 102681442A CN 2012101702375 A CN2012101702375 A CN 2012101702375A CN 201210170237 A CN201210170237 A CN 201210170237A CN 102681442 A CN102681442 A CN 102681442A
- Authority
- CN
- China
- Prior art keywords
- fault
- control
- tolerant
- intelligent
- model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Abstract
The invention discloses an intelligent fault-tolerant control system and a control method for zonal electric distribution of a full-electric boat. A full-electric boat power system is divided into a generator set control sub-system, an electric propulsion control sub-system, a power grid and a power distribution control sub-system of the power grid by a power distribution system; the control mode of each control sub-system is controlled by adopting an intelligent fault-tolerant control method of reference and model identification based on a neural network model; and the control method has a learning function and can be used for improving the control ability of each sub-system on the fault, so that the fault control ability of the full-electric boat power system is improved. With the adoption of a power supply sub-system and a power utilization sub-system of the designed boat power control system, flexible combination of power configuration of meeting navigation requirement of the boat can be achieved, and fault isolation and fault-tolerant control can be carried out. The power system structure and the fault-tolerant control method disclosed by the invention can adapt to safe and reliable requirements of the electric propulsion boat power system on the fault control.
Description
Technical field
The present invention relates to complete electric ship electric system zone distribution control technology, be specifically related to a kind of intelligent fault-tolerance control system and control corresponding method that is used for the regional distribution of the reliable control of complete electric ship.
Background technology
China is world shipping big country and shipbuilding big country, and the gordian technique of Ship Electrical Power System robotization aspect is in great demand.Boats and ships such as modern capital ship, engineering ship, pleasure boat are because at the high request of aspects such as ship manoeuverability, energy-conservation and environmental protection; Begin to adopt the power of electric propulsion as ship's navigation; Changed boats and ships and relied on the situation of diesel engine as thrust power; Produced the complete electrified revolution of ship power equipment and control thereof, having formed is the complete electrified boats and ships of boats and ships active force and energy core thereof with the electric system.
Electrical Propulsion Ship expectation set and constant speed navigation under oceanic winds/current uncertain environment.Ship Swaying is violent when bad weather, and the hydrodynamic force load variations of propelling screws is big, receives the impact of huge uncertainty load as the electric system of boats and ships active force, and the probability of fault strengthens.At present, the marine ship electric system is in the majority with the electric system of the electric system of dendritic tie lines form and the linear formula of ring junction, and the ability of system's control fault is relatively weaker; Also have some Ship Electrical Power Systems to adopt the form of regional distribution, mostly its control method is conventional control method, does not have the fault-tolerant control ability for the brute force of fault; Therefore, Ship Electrical Power System is lower for the control ability of fault at present, relies on hardware device redundancy and relay protection handling failure thereof mostly, and the electric system fluctuating range is big during fault.
Summary of the invention
The present invention is directed to spot ship electric system existing deficiency aspect the fault-tolerant control of fail-safe control, and a kind of regional distribution fault-tolerant control system of complete electric ship be provided---based on the regional distribution intelligent fault-tolerance control system and the control method of regional distribution system enforcement.The present invention is used to solve the control problem of complete electric ship electric power system fault, and the system that makes still can keep the continuous power supply to visual plant when breaking down, and the safe reliability of electric system is greatly improved.Adopt based on the intelligent multivariate fault tolerant control method of the neural network of model reference and Model Distinguish for the control of genset subsystem, electric propulsion subsystem, electrical network and the distribution subsystem thereof of complete electric ship.Through the fault-tolerant control system and the method for regional distribution, realize the safe and reliable control of complete electric ship electric system.
In order to achieve the above object, the present invention adopts following technical scheme:
A kind of complete electric ship zone distribution intelligent fault-tolerance control system, this control system matches with the regional distribution system of complete electric ship, and this control system comprises:
Fault-tolerant controller: the unit that is used to implement the intelligent fault-tolerance control algolithm; This controller is realized based on neural network model and control method thereof; Said controller serves as according to learning and revise, implementing fault-tolerant control with the deviation between the subsystem controlled device output of the regional distribution system of output of neural network reference model and complete electric ship; Fault-tolerant controller is accomplished the calculating of intelligent fault-tolerance control algolithm, exports fault-tolerant control execution command, to being sent control command by control subsystem, implements fail-safe control;
Control dispenser: make decisions with separating according to fault detect, carry out selectivity for the instruction of fault-tolerant controller and distribute, make control command confirm determine with the result who separates by fault detect;
Actuator: carry out the fault-tolerant steering order that control dispenser is selected distribution, change system control amount or system architecture;
Sensor and transmitter: the detection subsystem physical quantity, and convert the unified standard signal into, be convenient to transmission and computer acquisition;
Neural network (NN) generalized object model: with neural net method the subsystem actual motion course of work is learnt, be used to obtain the model of the remaining element summation of real system except that controller;
Neural network (NN) reference model: with neural net method subsystem is normally moved the course of work and carry out off-line learning, the model of the remaining element summation except that fault-tolerant controller of acquisition plays contrast standard when being used for the real system operation;
Intelligent trouble detects and separative element: data collection and analysis is carried out in the I/O to each subsystem Be Controlled object and sensor and transmitter in the regional distribution system of actuator, complete electric ship, the fault detect of using neural network intelligent trouble pattern-recognition carrying out subsystem with separate;
Fault-tolerant control decision unit: the information such as the fault type that provides with separative element, rank, regional extent that detect according to intelligent trouble are made a strategic decision, and confirm to adopt the mode of the fault-tolerant control of one of system reconfiguration, degraded running, Fault Compensation, failover or combination.
In the instantiation of fault-tolerant control system scheme; Said intelligent trouble detects with separative element and comprises actuator FDI, control system FDI, sensor transducer FDI, respectively actuator, RACS, sensor is accomplished fault detect and is separated with transmitter.
Further; Said fault-tolerant control decision unit comprises system reconfiguration module, degraded running module, Fault Compensation module, fault switch module, accomplishes the calculating of the fault-tolerant control algolithm of system reconfiguration, the calculating of the fault-tolerant control algolithm of degraded running, the calculating of the fault-tolerant control algolithm of Fault Compensation, the calculating of the fault-tolerant control algolithm of failover respectively.
As second portion of the present invention, a kind of complete electric ship zone distribution intelligent fault-tolerance control method, this method comprises the steps:
1, through path control
Said through path control comprises control of reference model path and the control of intelligent fault-tolerance control path, and said reference model path control procedure is following:
The expectation control signal is transmitted through neural network (NN) reference model; The output back provides standard information for the operation of controlled device; The output of controlled device and this standard information compare, and the deviation of generation reaches fault-tolerant controller and learns, and change control algolithm;
Said intelligent fault-tolerance control path control procedure is following:
Fault-tolerant controller is implemented algorithm computation, the generation steering order of fault-tolerant control; The instruction that produces is carried out through actuator, and the RACS controlled device is accomplished various operating modes; Neural network (NN) generalized object model is accomplished the online generalized model identification of controlled device system simultaneously; Its output compares with the output of RACS controlled device; The deviation that produces is used for the study of Model Distinguish unit; Make the actual ten minutes of the RACS plant model that obtains and object approaching, failure condition that can reflection object during fault;
2, backfeed loop fault diagnosis
The input and output of actuator, RACS controlled device and sensor and transmitter are input to intelligent trouble through sense channel and detect and separative element, implement fault detect and separate through Intelligent Information Processing;
Fault-tolerant control decision unit detects with separative element acquisition fault detect from intelligent trouble and makes a strategic decision with the concluding information of separating, and provides the resolution policy of fault-tolerant controller selective after the decision-making;
3, Model Distinguish process
State under the various abundant excitation operating mode of neural network (NN) reference model off-line identification system when normal operation is accomplished the Model Distinguish to system's normal condition, reflects the running status that the intelligent fault-tolerance control system is desirable with this.The output of neural network (NN) reference model and RACS controlled device compares the generation deviation, and with the learning process of this drift correction fault-tolerant controller;
Neural network (NN) generalized object model is realized the state under the various operating conditions of online identification system; Completion is to the Model Distinguish of system's real time execution; The real-time running state of reflection system; Various information when model has comprised the system failure; Output through its output and RACS controlled device relatively produces deviation, and deviation signal is used for the correction of neural network (NN) generalized object model, and the neural network of acquisition (NN) generalized object model algorithm is the important composition of the algorithm of intelligent fault-tolerance controller.
The present invention reliably controls requirement for adapting to complete electric ship power system security, is the subsystem of three types of diesel generator set, electric propulsion motor and system thereof, electrical network and the distribution systems thereof of complete electric ship with system decomposition.Diesel generator set, electric propulsion electric system utilization neural network intelligent fault-tolerance control method are implemented control, improve reliability; On this basis, complete electric ship electrical network and distribution system thereof implement to combine comprehensive coordination and the electrical network intelligent fault-tolerance control of protecting.
System architecture provided by the invention and control method take into full account the requirement of the safe navigation of boats and ships ocean freestanding environment; Be convenient to control the fault of sensor, actuator and the electrical equipment generation of diesel generator set, electric propulsion motor and system, electrical network and the distribution system thereof of complete electric ship; The safe reliability of complete electric ship electric system is improved; Still can keep safe operation when complete electric ship electric system is broken down, have fault-tolerant control for fault.
Description of drawings
Further specify the present invention below in conjunction with accompanying drawing and embodiment.
Fig. 1 is the structural drawing of complete electric ship electric system of the present invention zone distribution system;
Fig. 2 is the block scheme of the intelligent fault-tolerance control system based on neural network of the present invention.
Embodiment
For technological means, creation characteristic that the present invention is realized, reach purpose and effect and be easy to understand and understand, below in conjunction with concrete diagram, further set forth the present invention.
Complete electric shipping with electric propulsion, with Ship Electrical Power System as boats and ships main power systems, the requirement that electric system has been proposed to keep high power supply safety and reliability.The core system safe and reliable operation that electric system is supplied with as complete electric ship energy is very crucial; Automated system is long-time, operation can be broken down inevitably continuously, and therefore, complete electric ship electric system is very important for the control of fault.Adopting the regional distribution method of electric system, can the electric system of complicacy be decomposed into the combination of some functional subsystems, can be the method for the complete electric ship reliability of a kind of effective raising according to the different isolated faults in time in zone.Therefore,, the method ACTIVE CONTROL fault of fault-tolerant control is used in the automatic control of complete electric ship electric system, improved safe reliability and have great importance improving the continuous power supply of boats and ships qualitatively.
Thus; The present invention is the functional subsystem that complete electric ship electric system is divided into three types of genset subsystem, electric propulsion subsystem, electrical network and distribution subsystem thereof with complete electric ship electric system various piece by Function Decomposition, requires the subregion to dispose whole Ship Electrical Power System according to safety of ship and fail-safe control then.
The genset subsystem 11,12,21,22 of Fig. 1; For whole complete electric ship electric system provides electric power; Two or above diesel generator set 7,8,9,10,17,18,19,20 are arranged in each genset subsystem; Redundant each other and additional, strengthen fault-tolerant ability for fault, the control of every genset is undertaken by intelligent fault-tolerance control structure and the method for Fig. 2.
The electric propulsion subsystem 3,5,23 of Fig. 1 provides the power of ship's navigation, and each electric propulsion subsystem has a propulsion electric machine 2,6,24 and propeller propulsion system 1,4,25 thereof.The control of such subsystem is undertaken by the intelligent fault-tolerance control method of Fig. 2 equally; Two ship stern electric propulsion subsystems of large ship configuration (Fig. 13,5) and stem electric propulsion subsystem (Fig. 1 23).
All the other are that General Electric (for example loads subsystem; 13,14,15,16); Be divided into several boats and ships loads and can isolate the electricity consumption zone, concrete quantity is divided configuration with the foundation in zone according to the importance security needs of boats and ships power load, electrical equipment; For example, boats and ships navigation electric/electronic device system region, service electrical system district, cabin, deck electrical system district, crew quarters, various passenger's service areas etc. can be set up the electrical load zone of independent or several districts associating.
Electrical network and distribution subsystem thereof are the summations of whole controller switching equipments and electrical network and protective device thereof, comprise the distribution of genset subsystem, distribution and the electrical network and the protective device thereof of electric propulsion subsystem; Its opertaing device is distributed among genset subsystem (Fig. 1 11,12,21,22), electric propulsion subsystem (Fig. 13,5,23), universal electric power load subsystem (Fig. 1 13,14,15, the 16) functional module.
The topological structure of the electrical network flexible combination of complete in the present invention electric ship electric system meets the needs of ship's navigation power, and is following in conjunction with its fault-tolerant control operational process of Fig. 1:
(1) genset subsystem 11,12,21,22 can be implemented power supply to any electric propulsion subsystem 3,5,23 and universal electric power load subsystem 13,14,15,16 through supply line arbitrarily.For example, the genset subsystem 11 of Fig. 1 can be respectively through 30 pairs of electric propulsion subsystems 3 of supply line, through 32 and 31 pairs of electric propulsion subsystems 5, supply power through 27 and 38 pairs of electric propulsion subsystems 23.
(2) for the power supply of electric propulsion subsystem 3,5,23, two of needs and being incorporated into the power networks of above genset subsystem are supplied power generally speaking, and case of emergency allows single generator group subsystem that the electric propulsion subsystem is supplied power down.For example, 3 of Fig. 1 can be incorporated into the power networks through supply line 26 through supply line 30 and 21 by 11 and supply power; Also can increase by 12,22 again and be incorporated into the power networks, participate in power supply; Case of emergency allows 11 to supply power to 3 through monomotor systems such as 26 through 30 or 21.
(3) universal electric power load subsystem then allows any single generator group subsystem to supply power, and being incorporated into the power networks of one or more genset subsystems can be supplied power to one or more universal electric power load subsystems.For example, single generator group subsystem powering mode: 11 supply power through 33 pairs 13, and 11 supply power, supply power, supply power through 42 pairs 16 through 41 pairs 15 again through 33 and 40 pairs 14; The multiple generator group subsystem powering mode that is incorporated into the power networks: 21 make up through 27 and 11, and 12 make up through 32 and 11, form 2 or 3 genset subsystems powering mode that is incorporated into the power networks, and implement power supply.
Therefore; The topological structure of electric of native system is fit to the power supply combination of different genset subsystems for different power load subsystems; When any genset subsystem broke down, other genset subsystem can substitute fault generator group subsystem and implement power supply; When any power load system breaks down, can isolate this power load system, do not influence the power supply of other power load; The electric propulsion subsystem of ship stern has subsequent use each other function each other.Therefore, this system's electric network composition has formed the topological structure form of system's power supply and the flexible combination that receives electricity, when satisfying ship's navigation for the requirement of power configuration.Whole complete electric ship electric system distributes through subsystem and the modular design of subsystem function realizes having the fault tolerance for fault.
Based on the regional distribution system of above-mentioned complete electric ship, the invention provides based on the neural network model reference and the intelligent fault-tolerance control method of Model Distinguish each subsystem is implemented fault-tolerant control.
Subsystem for three types of diesel generator set, electric propulsion motor and the system thereof of complete electric ship, electrical network and distribution systems thereof; Control is implemented in the control of utilization neural network intelligent fault-tolerance respectively; Having formed with the automatic system of marine diesel-generator pack module is that the intelligent fault-tolerance RACS of object, the system module that constitutes with marine vessel power propulsion electric machine and screw propeller thereof are the intelligent fault-tolerance RACS of object, are the intelligent fault-tolerance RACS of object with network of ship and distribution system module thereof, and their basic control mode and program are mutually roughly the same.
Make subsystem controlled when breaking down through corresponding intelligent fault-tolerance RACS, system still can keep continuous operation, makes Ship Electrical Power System keep power supply continuously.As shown in Figure 2 based on the neural network model reference with the structure of the intelligent fault-tolerance control system of Model Distinguish.
This control system comprises:
Fault-tolerant controller 2, control dispenser 3, actuator 4, sensor and transmitter 6, neural network (NN) generalized object model 13, neural network (NN) reference model 7, intelligent trouble detect and separative element 22, fault-tolerant control decision unit 21.
Wherein, Fault-tolerant controller: the unit that is used to implement the intelligent fault-tolerance control algolithm; This controller is realized based on neural network model and control method thereof; Said controller serves as according to learning and revise, implementing fault-tolerant control with the deviation between the subsystem controlled device output of the regional distribution system of output of neural network reference model and complete electric ship; Fault-tolerant controller is accomplished the calculating of intelligent fault-tolerance control algolithm, exports fault-tolerant control execution command, to being sent control command by control subsystem, implements fail-safe control;
Control dispenser: make decisions with separating according to fault detect, carry out selectivity for the instruction of fault-tolerant controller and distribute, make control command confirm determine with the result who separates by fault detect;
Actuator: carry out the fault-tolerant steering order that control dispenser is selected distribution, change system control amount or system architecture;
Sensor and transmitter: the detection subsystem physical quantity, and convert the unified standard signal into, be convenient to transmission and computer acquisition;
Neural network (NN) generalized object model: with neural net method the subsystem actual motion course of work is learnt, be used to obtain the model of the remaining element summation of real system except that controller;
Neural network (NN) reference model: with neural net method subsystem is normally moved the course of work and carry out off-line learning, the model of the remaining element summation except that fault-tolerant controller of acquisition plays contrast standard when being used for the real system operation;
Intelligent trouble detects and separative element: data collection and analysis is carried out in the I/O to each subsystem Be Controlled object and sensor and transmitter in the regional distribution system of actuator, complete electric ship, the fault detect of using neural network intelligent trouble pattern-recognition carrying out subsystem with separate;
Intelligent trouble detects with separative element and comprises actuator FDI (23), control system FDI (24), sensor transducer FDI (25); Wherein FDI is fault detect and separate, and these three unit are accomplished fault detects to actuator 4, control system, sensor with transmitter 6 respectively and separated.
In the present invention, the FDI intelligent trouble detects with the method for separating the identification of utilization network mode and carries out, and neural network is at first learnt a large amount of fault cases, obtains the mapping relations between fault and the sign, for fault identification lays the first stone; The neural network that to succeed in school then is used for online fault detect, and the data that collect are carried out real-time operation, discerns out of order characteristic, confirms the source of trouble.
Fault-tolerant control decision unit: the information such as the fault type that provides with separative element, rank, regional extent that detect according to intelligent trouble are made a strategic decision, and confirm the fault-tolerant control mode with system reconfiguration, degraded running, Fault Compensation, failover.
Fault-tolerant control decision unit comprises system reconfiguration module, degraded running module, Fault Compensation module, fault switch module, accomplishes the calculating of the fault-tolerant control algolithm of system reconfiguration, the calculating of the fault-tolerant control algolithm of degraded running, the calculating of the fault-tolerant control algolithm of Fault Compensation, the calculating of the fault-tolerant control algolithm of failover respectively.
Generally, each controlled device among Fig. 1 all adopts the control forms of Fig. 2 to implement intelligent fault-tolerance control.In conjunction with Fig. 2 and Fig. 1, explain that relation each other is following.
Among Fig. 25 is controlled devices of control; As the genset RACS; 5 of Fig. 2 can be 7 among Fig. 1,8,9,10,17,18,19, one of 20, i.e. 7,8,9,10,17,18,19,20 control is to adopt Fig. 2 structure to carry out fault-tolerant control.As the propulsion electric machine RACS, 5 of Fig. 2 can be 2 among Fig. 1,6, one of 24, i.e. 2,6,24 control is to adopt Fig. 2 structure to carry out fault-tolerant control.As the power grid control subsystem, 5 of Fig. 2 is power supply network combinations of 26 to 45, and promptly the power supply network of 26 to 45 compositions is to adopt Fig. 2 structure to carry out fault-tolerant control.They for the Different control subsystem, all be to adopt the control structure of Fig. 2 to carry out fault-tolerant control, but the content of each unit is different to liking according to Different control in the structure.W among Fig. 2 is the expectation vector of control, and mainly the electric parameter desired amount by electric system combines; Be in 5 position of control loop by control subsystem, detect its output signal through 6.Among Fig. 22,3,7,13,21,22 forms the intelligent fault-tolerance control framework; 2,3 as the major control device, and 7 provide the The whole control system comparative standard of controlling schemes, the state of 12 on-line identification objects; 22 pairs 4,5,6 I/O gather and analyze, and implement the neural network intelligent trouble and detect and separate; The result of 21 pairs of fault detects judges, fault-tolerant control strategy is provided, and offers 3 through 18 and selects, and carries out the scheme implementation that intelligent fault-tolerance is controlled by 2.
Different from control subsystem; Among Fig. 22,3,5,7,13,21,22 content are different, for example, and for the diesel-driven generator RACS; These unit all are the corresponding units in the generator control system; The 5th, diesel-driven generator, the 7th, the neural network reference model of the diesel-driven generator RACS that off-line identification obtains, the 13rd, the model of the gen-set control system of real-time online identification.
The present invention is directed to the real system fault and have uncertainty at random, electrical equipments such as the actuator of system, sensor and transmitter, controller, controlled device all might break down.Intelligent fault-tolerance control system provided by the invention comprises neural network (NN) reference model, neural network generalized object model, the fault-tolerant decision-making of neural network, neural network fault-tolerant controller and five unit of control dispenser thereof; The fault-tolerant decision-making of neural network comprises decision-making modes such as system's on-line reorganization, degraded running, Fault Compensation and failover; Select neuron algorithm, learning rules, the network topology structure of neural network fault-tolerant controller to carry out the control law intelligent reconstruction through control dispenser, implement fault-tolerant control.The controlled target of intelligent fault-tolerance control is: make controlled device and each unit thereof when certain fault occurring, system still has the stable operation characteristics approaching with the neural network reference model.The control procedure of intelligent fault-tolerance control is implemented the program of fault-tolerant control then and is carried out according at first carrying out fault detect, diagnose and separating.The neural network reference model obtains through off-line identification controlled device in advance; Neural network obtains the generalized object model through on-line study; Model information comprises in the system except that controller and the whole remaining environment division the reference model; Be used to discern the running status and the failure condition of generalized object; Produce the Back ground Information of controlled device fault diagnosis, offer the fault handling neural network and accomplish system failure detection, diagnose and separate.The quick fault testing of neural network, diagnose with separating and comprise the extraction of fault signature value, fault detect, failure modes and separate five processes with identification, fault diagnosis, intelligent fault.Intelligent fault detects and separates and can carry out fault diagnosis and separating treatment respectively to actuator, sensor and transmitter, control system, also can totally carry out fault diagnosis and separating treatment to each subsystem.
Mainly accomplish the fault-tolerant control of genset and sensor thereof, actuator, controller failure for the control of complete electric ship diesel generator set; Control the fault-tolerant control of mainly accomplishing propulsion electric machine and sensor thereof, actuator, controller failure for complete electric ship electric propulsion motor and system thereof; For complete electric ship electrical network and the main fault-tolerant control of accomplishing regional power grid and controller switching equipment fault of distribution system control thereof, the comprehensive coordination and protection control of complete electric ship electric system carried out in complete electric ship electrical network and distribution system control thereof simultaneously.
Based on above-mentioned intelligent fault-tolerance control system; The present invention also provides corresponding intelligent fault-tolerance control method; This control method is the basic control method that is fit to diesel generator set, electric propulsion motor and system thereof, electrical network and three kinds of subsystems of distribution system thereof of complete electric ship, in order to various subsystems are realized intelligent fault-tolerance control.In the complete electric ship electric system; Diesel generator set is as feeder ear; Electric propulsion motor and system thereof are as load end, and electrical network and distribution system thereof couple together them, and therefore the fault-tolerant control of complete electric ship electric system is on the fault-tolerant control basis of three types of subsystems, to realize; Possessed the function of fault-tolerant control when each subsystem after, the fault-tolerant control function of complete electric ship electric system just can be guaranteed.
In conjunction with accompanying drawing 2, fault-tolerant controlling schemes realizes through through path control, backfeed loop fault diagnosis and three control procedures of Model Distinguish.Through path control, backfeed loop fault diagnosis and three control procedures of Model Distinguish, realized for Fig. 2 " neural network failure of the data acquisition of Unit 2,4,5,6 → 22 detect with separate → Control Allocation of 21 fault-tolerant decision-making → 3 selects the instruction of → 2 enforcement intelligent fault-tolerance control → 4 to carry out → act on 5 controlled devices " the closed-loop control course of work.
The implementation process explanation of fault-tolerant control action as follows.
1. through path control procedure
Through path is the major control passage of system, is divided into two passages in reference model path and intelligent fault-tolerance control path.
(1) reference model path
The expectation control signal is transmitted through 7 of Fig. 2, and output back provides standard information for the operation of controlled device, and the output of controlled device 5 and this standard compare, the deviation of generation through 8 be used for intelligent controller 2 study, the change control algolithm; Be the unit that embodies the learning functionality of control method,, make control algolithm adapt to different controlled devices through study.
(2) intelligent fault-tolerance control path
The 2,4, the 5th of Fig. 2, the major equipment of system, the flow process by arrow of its signal has been formed the intelligent fault-tolerance control path.2 of Fig. 2 is fault-tolerant controllers, and algorithm computation, the instruction of implementing fault-tolerant control produce; The instruction that produces is carried out through 4, and various operating modes are accomplished in control 5.What be worth to propose is 2,4,5 in operational process, all might break down, and the safe reliability of system is descended.12 accomplish the generalized model identification of controlled device systems, and its output compares with 5 output, and the deviation of generation is used for the study of Model Distinguish unit, make the actual ten minutes of 12 models that obtain and object approaching, the failure condition of reflection object during fault; Be another unit that embodies the learning functionality of control method,, make control algolithm adapt to different controlled devices through study.12 identification result directly influences 2 algorithm formation, is the essential elements that control algolithm is formed.
2. backfeed loop failure diagnostic process
The data acquisition that backfeed loop is accomplished information and fault detect with separate.6 of Fig. 2 is sensor and transmitter unit, detects the electric parameter of controlled device 5, converts standard signal to and is input to control module, accomplishes the data acquisition task.6 might break down in operational process, and the safe reliability of system is descended.The 22nd, intelligent trouble detects and separative element, and 4,5,6 input and output are input to 22 through 14,15,16,19,20 sense channel, implements fault detect and separates through Intelligent Information Processing.The 21st, fault-tolerant control decision unit from 22 acquisition fault detects and the concluding information of separating, carries out classification by 21 and makes a strategic decision, and provides the resolution policy of fault-tolerant control selective after the decision-making.22 23,24,25 respectively mapping carry out the fault detect of performance element, control system, sensor unit and the information processing that separates, 23,24,25 algorithm mapping designs in different unit, content is different.
The basic controlling process of intelligent fault-tolerance control method interosculates for the ultimate principle of diesel-driven generator subsystem, electric propulsion RACS; The intelligence of control is embodied in the learning functionality that control method has a plurality of unit.For the intelligent fault-tolerance control of universal electric power load subsystem, object can relate to marine vessel power power supply network and distribution thereof, and the control of load can relate to the relay protection control content, and the content of fault-tolerant control will comprise these contents.Intelligent fault-tolerance control is primarily aimed at the fault of Unit 2,4,5,6 of Fig. 2 and implements intelligent fault-tolerance control.
3. Model Distinguish process
The identification of system model is mainly accomplished by Unit 7,13 of Fig. 2, and the 7th, the standard reference model of system is that the neural network off-line carries out identification and the model of learning to produce to RACS integral body.State under main identification, the various abundant excitation operating mode of learning system when normal operation; Completion is to the Model Distinguish of system's normal condition; Reflected the desirable running status of intelligent fault-tolerance control system; 7 effect provides the standard signal of system's operation during working control, makes the output of the output of system near this master pattern; 7 with the deviation of controlled device output through 8 revise fault-tolerant controllers learning process.The 13rd, system generalized object Model Distinguish unit is that Neural Network Online is carried out identification and the model of learning to produce to RACS, and 13 outputs through its output and 5 relatively produce deviation, and deviation signal is through 12 modification model algorithms.13 effect mainly is the state under on-line identification, the various operating conditions of learning system, accomplishes the Model Distinguish to system's real time execution, has reflected the real-time running state of system, the various information when having comprised the system failure; Use with the deviation of controlled device output and proofread and correct 13 algorithm, the model that makes acquisition is with actual approaching, and the algorithm that the algorithm of 13 acquisitions is used for intelligent fault-tolerance controller 2 forms.The system model identification is the key component that intelligent fault-tolerance controlling schemes control algolithm forms, and directly influences the quality of system's control.
In this method, the neural network reference model obtains through off-line identification controlled device in advance; Neural network obtains the generalized object model through on-line study; Model information comprises in the system except that controller and the whole remaining environment division the reference model; Be used to discern the running status and the failure condition of generalized object; Produce the Back ground Information of controlled device fault diagnosis, offer the fault handling neural network and accomplish system failure detection, diagnose and separate.
The quick fault testing of neural network, diagnose with separating and comprise the extraction of fault signature value, fault detect, failure modes and separate five processes with identification, fault diagnosis, intelligent fault.
Intelligent fault detects and separates and can carry out fault diagnosis and separating treatment respectively to actuator, sensor and transmitter, control system, also can totally carry out fault diagnosis and separating treatment to each subsystem.
Mainly accomplish the fault-tolerant control of genset and sensor thereof, actuator, controller failure for the control of complete electric ship diesel generator set; Control the fault-tolerant control of mainly accomplishing propulsion electric machine and sensor thereof, actuator, controller failure for complete electric ship electric propulsion motor and system thereof; For complete electric ship electrical network and the main fault-tolerant control of accomplishing regional power grid and controller switching equipment fault of distribution system control thereof, the comprehensive coordination and protection control of complete electric ship electric system carried out in complete electric ship electrical network and distribution system control thereof simultaneously.
More than show and described ultimate principle of the present invention, principal character and advantage of the present invention.The technician of the industry should understand; The present invention is not restricted to the described embodiments; That describes in the foregoing description and the instructions just explains principle of the present invention; Under the prerequisite that does not break away from spirit and scope of the invention, the present invention also has various changes and modifications, and these variations and improvement all fall in the scope of the invention that requires protection.The present invention requires protection domain to be defined by appending claims and equivalent thereof.
Claims (4)
1. complete electric ship zone distribution intelligent fault-tolerance control system, this control system matches with the regional distribution system of complete electric ship, it is characterized in that this control system comprises:
Fault-tolerant controller: the unit that is used to implement the intelligent fault-tolerance control algolithm; This controller is realized based on neural network model and control method thereof; Said controller serves as according to learning and revise, implementing fault-tolerant control with the deviation between the subsystem controlled device output of the regional distribution system of output of neural network reference model and complete electric ship; Fault-tolerant controller is accomplished the calculating of intelligent fault-tolerance control algolithm, exports fault-tolerant control execution command, to being sent control command by control subsystem, implements fail-safe control;
Control dispenser: make decisions with separating according to fault detect, carry out selectivity for the instruction of fault-tolerant controller and distribute, make control command confirm determine with the result who separates by fault detect;
Actuator: carry out the fault-tolerant steering order that control dispenser is selected distribution, change system control amount or system architecture;
Sensor and transmitter: the detection subsystem physical quantity, and convert the unified standard signal into, be convenient to transmission and computer acquisition;
Neural network generalized object model: with neural net method the subsystem actual motion course of work is learnt, be used to obtain the model of the remaining element summation of real system except that controller;
The neural network reference model: with neural net method subsystem is normally moved the course of work and carry out off-line learning, the model of the remaining element summation except that fault-tolerant controller of acquisition plays contrast standard when being used for the real system operation;
Intelligent trouble detects and separative element: data collection and analysis is carried out in the I/O to each subsystem Be Controlled object and sensor and transmitter in the regional distribution system of actuator, complete electric ship, the fault detect of using neural network intelligent trouble pattern-recognition carrying out subsystem with separate;
Fault-tolerant control decision unit: the information such as the fault type that provides with separative element, rank, regional extent that detect according to intelligent trouble are made a strategic decision, and confirm to adopt the mode of the fault-tolerant control of one of system reconfiguration, degraded running, Fault Compensation, failover or combination.
2. a kind of complete electric ship according to claim 1 zone distribution intelligent fault-tolerance control system; It is characterized in that; Said intelligent trouble detects with separative element and comprises actuator FDI, control system FDI, sensor transducer FDI, respectively actuator, RACS, sensor is accomplished fault detect and is separated with transmitter.
3. a kind of complete electric ship according to claim 1 zone distribution intelligent fault-tolerance control system; It is characterized in that; Said fault-tolerant control decision unit comprises system reconfiguration module, degraded running module, Fault Compensation module, fault switch module, accomplishes the calculating of the fault-tolerant control algolithm of system reconfiguration, the calculating of the fault-tolerant control algolithm of degraded running, the calculating of the fault-tolerant control algolithm of Fault Compensation, the calculating of the fault-tolerant control algolithm of failover respectively.
4. a complete electric ship zone distribution intelligent fault-tolerance control method is characterized in that said method comprises the steps:
(1) through path control
Said through path control comprises control of reference model path and the control of intelligent fault-tolerance control path, and said reference model path control procedure is following:
The expectation control signal is transmitted through the neural network reference model, and the output back provides standard information for the operation of controlled device, and the output of controlled device and this standard information compare, and the deviation of generation reaches fault-tolerant controller and learns, and changes control algolithm;
Said intelligent fault-tolerance control path control procedure is following:
Fault-tolerant controller is implemented algorithm computation, the generation steering order of fault-tolerant control; The instruction that produces is carried out through actuator, and the RACS controlled device is accomplished various operating modes; Neural network generalized object model is accomplished the online generalized model identification of controlled device system simultaneously; Its output compares with the output of RACS controlled device; The deviation that produces is used for the study of Model Distinguish unit; Make the actual ten minutes of the RACS plant model that obtains and object approaching, failure condition that can reflection object during fault;
(2) backfeed loop fault diagnosis
The input and output of actuator, RACS controlled device and sensor and transmitter are input to intelligent trouble through sense channel and detect and separative element, implement fault detect and separate through Intelligent Information Processing;
Fault-tolerant control decision unit detects with separative element acquisition fault detect from intelligent trouble and makes a strategic decision with the concluding information of separating, and provides the resolution policy of fault-tolerant controller selective after the decision-making;
(3) Model Distinguish process
State under the various abundant excitation operating mode of neural network reference model off-line identification system when normal operation is accomplished the Model Distinguish to system's normal condition, reflects the running status that the intelligent fault-tolerance control system is desirable with this.The output of neural network reference model and RACS controlled device compares the generation deviation, and with the learning process of this drift correction fault-tolerant controller;
Neural network generalized object model is realized the state under the various operating conditions of online identification system; Completion is to the Model Distinguish of system's real time execution; Various information when the real-time running state of reflection system, model have comprised the system failure relatively produce deviation through its output with the output of RACS controlled device; Deviation signal is used for the correction of neural network generalized object model, and the neural network generalized object model algorithm of acquisition is the important composition of intelligent fault-tolerance controller algorithm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210170237.5A CN102681442B (en) | 2012-05-28 | 2012-05-28 | Intelligent fault-tolerant control system and control method for zonal power distribution of full-electric boat |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210170237.5A CN102681442B (en) | 2012-05-28 | 2012-05-28 | Intelligent fault-tolerant control system and control method for zonal power distribution of full-electric boat |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102681442A true CN102681442A (en) | 2012-09-19 |
| CN102681442B CN102681442B (en) | 2014-08-20 |
Family
ID=46813519
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210170237.5A Expired - Fee Related CN102681442B (en) | 2012-05-28 | 2012-05-28 | Intelligent fault-tolerant control system and control method for zonal power distribution of full-electric boat |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102681442B (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103353752A (en) * | 2013-04-12 | 2013-10-16 | 北京航空航天大学 | Aircraft environment control system control component fault diagnosis method based on four-level RBF neural network |
| CN105204335A (en) * | 2015-09-15 | 2015-12-30 | 南京航空航天大学 | Fault-tolerant control method and system of under-actuated rigid body system |
| CN107491848A (en) * | 2017-09-29 | 2017-12-19 | 集美大学 | Dynamic positioning ship energy management optimization method with automatic fault isolation |
| CN107818205A (en) * | 2017-10-23 | 2018-03-20 | 中国船舶工业集团公司第七0八研究所 | The power supply trouble model and impact analysis method of a kind of naval vessel visual plant |
| CN107949812A (en) * | 2015-06-29 | 2018-04-20 | 苏伊士集团 | For detecting the abnormal combined method in water distribution system |
| CN109407635A (en) * | 2018-11-06 | 2019-03-01 | 哈尔滨工程大学 | A kind of ship craft integrated electric propulsion safety control system smart system structure |
| CN109800227A (en) * | 2018-12-26 | 2019-05-24 | 贵州电网有限责任公司 | Dispatching of power netwoks utilization logger system based on Structured Design |
| CN109981339A (en) * | 2019-02-14 | 2019-07-05 | 电子科技大学 | A kind of failure tolerant control method of net connection automotive control system |
| CN112947374A (en) * | 2021-02-09 | 2021-06-11 | 上海海事大学 | Intelligent self-healing control method for electric propulsion of regional distribution ship |
| CN113110464A (en) * | 2021-04-22 | 2021-07-13 | 集美大学 | Intelligent full-electric ship path planning method capable of reducing energy consumption |
| CN113562065A (en) * | 2021-07-15 | 2021-10-29 | 东风汽车集团股份有限公司 | Control method and device for preventing data abnormality of electric power steering controller |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5048100A (en) * | 1988-12-15 | 1991-09-10 | Michael Kuperstein | Self organizing neural network method and system for general classification of patterns |
| CN101783563A (en) * | 2010-03-10 | 2010-07-21 | 江苏大学 | Multiple multiphase fault-tolerant motors for electric automobile and decoupling control method thereof |
| CN101995821A (en) * | 2010-06-25 | 2011-03-30 | 哈尔滨工程大学 | Intelligent stepping fault-tolerance control method |
-
2012
- 2012-05-28 CN CN201210170237.5A patent/CN102681442B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5048100A (en) * | 1988-12-15 | 1991-09-10 | Michael Kuperstein | Self organizing neural network method and system for general classification of patterns |
| CN101783563A (en) * | 2010-03-10 | 2010-07-21 | 江苏大学 | Multiple multiphase fault-tolerant motors for electric automobile and decoupling control method thereof |
| CN101995821A (en) * | 2010-06-25 | 2011-03-30 | 哈尔滨工程大学 | Intelligent stepping fault-tolerance control method |
Non-Patent Citations (3)
| Title |
|---|
| JOSE R. NORIEGA，ET AL.: "A Direct Adaptive Neural-Network Control for Unknown Nonlinear Systems and Its Application", 《IEEE TRANSACTIONS ON NEURAL NETWORKS》 * |
| 常锋等: "基于神经网络的模型参考飞控系统容错控制", 《控制工程》 * |
| 施伟锋: "船舶柴油发电机转速人工神经网络控制", 《上海海事大学学报》 * |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103353752B (en) * | 2013-04-12 | 2015-10-28 | 北京航空航天大学 | Based on the aircraft environmental control system Control Component method for diagnosing faults of level Four RBF neural |
| CN103353752A (en) * | 2013-04-12 | 2013-10-16 | 北京航空航天大学 | Aircraft environment control system control component fault diagnosis method based on four-level RBF neural network |
| CN107949812B (en) * | 2015-06-29 | 2021-09-14 | 苏伊士集团 | Method for detecting anomalies in a water distribution system |
| CN107949812A (en) * | 2015-06-29 | 2018-04-20 | 苏伊士集团 | For detecting the abnormal combined method in water distribution system |
| CN105204335A (en) * | 2015-09-15 | 2015-12-30 | 南京航空航天大学 | Fault-tolerant control method and system of under-actuated rigid body system |
| CN107491848A (en) * | 2017-09-29 | 2017-12-19 | 集美大学 | Dynamic positioning ship energy management optimization method with automatic fault isolation |
| CN107818205A (en) * | 2017-10-23 | 2018-03-20 | 中国船舶工业集团公司第七0八研究所 | The power supply trouble model and impact analysis method of a kind of naval vessel visual plant |
| CN109407635A (en) * | 2018-11-06 | 2019-03-01 | 哈尔滨工程大学 | A kind of ship craft integrated electric propulsion safety control system smart system structure |
| CN109800227A (en) * | 2018-12-26 | 2019-05-24 | 贵州电网有限责任公司 | Dispatching of power netwoks utilization logger system based on Structured Design |
| CN109800227B (en) * | 2018-12-26 | 2022-04-22 | 贵州电网有限责任公司 | Power grid dispatching operation recording system based on structural design |
| CN109981339A (en) * | 2019-02-14 | 2019-07-05 | 电子科技大学 | A kind of failure tolerant control method of net connection automotive control system |
| CN109981339B (en) * | 2019-02-14 | 2021-03-30 | 电子科技大学 | Fault-tolerant control method of networked automobile control system |
| CN112947374A (en) * | 2021-02-09 | 2021-06-11 | 上海海事大学 | Intelligent self-healing control method for electric propulsion of regional distribution ship |
| CN113110464A (en) * | 2021-04-22 | 2021-07-13 | 集美大学 | Intelligent full-electric ship path planning method capable of reducing energy consumption |
| CN113562065A (en) * | 2021-07-15 | 2021-10-29 | 东风汽车集团股份有限公司 | Control method and device for preventing data abnormality of electric power steering controller |
| CN113562065B (en) * | 2021-07-15 | 2022-06-03 | 东风汽车集团股份有限公司 | Control method and device for preventing data abnormality of electric power steering controller |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102681442B (en) | 2014-08-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102681442B (en) | Intelligent fault-tolerant control system and control method for zonal power distribution of full-electric boat | |
| KR101498738B1 (en) | A method and a system for testing of a power management system of a marine vessel | |
| CN103869781B (en) | A kind of non-similar triplex redundance machine on-board electrical Electrical Load Management Center | |
| Jin et al. | Maritime DC microgrids-a combination of microgrid technologies and maritime onboard power system for future ships | |
| Shen et al. | Power and energy management in integrated power system | |
| CN106950962A (en) | A kind of unmanned boat dcs | |
| EP2330030A2 (en) | Device for controlling a vehicle | |
| CN111007852A (en) | System architecture of ship and intelligent ship | |
| CN105045105A (en) | State-delay four-rotor helicopter fault tolerant control device and method | |
| CN201576199U (en) | Ship engine room monitoring and alarming system based on CAN bus | |
| CN114217519A (en) | Self-reconfigurable marine network control system architecture and self-reconfigurable method | |
| CN112733962A (en) | Intelligent fault diagnosis and reconstruction method for ship regional power distribution power system | |
| CN104238546A (en) | Direct current master device fault diagnosis method based on hybrid neural network | |
| Yang et al. | Supervisory fault tolerant control with integrated fault detection and isolation: A switched system approach | |
| CN201576198U (en) | Ship cabin data observing and controlling system based on OPC technology | |
| CN113408941B (en) | Big data integration platform serving intelligent ship | |
| CN106414232B (en) | The mixed propulsion system and its control method of ship | |
| CN105334747A (en) | Ship dynamic positioning three-redundancy computer data voting synchronization method | |
| CN105305610A (en) | Ship electric propulsion monitoring device | |
| EP3321167A1 (en) | System for operating a boat equipped with an electric motor | |
| Saushev et al. | Rapid identification of the technical condition of a marine electric power system | |
| CN103246266B (en) | Industrial online maintenance-free control system | |
| CN104265567B (en) | The yaw system of band deviation wind direction protection automatically and control method thereof | |
| CN112865303A (en) | Self-sensing and self-diagnosing intelligent self-healing method for ship regional power distribution power system | |
| CN204424901U (en) | Drill ship DP3 closed loop PMS electric propulsion system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20140820 Termination date: 20210528 |