CN112034774A

CN112034774A - Hot redundancy control method

Info

Publication number: CN112034774A
Application number: CN202010696226.5A
Authority: CN
Inventors: 朱煌庆; 林尚飞; 孟令卫
Original assignee: 715th Research Institute of CSIC
Current assignee: 715th Research Institute of CSIC
Priority date: 2020-07-20
Filing date: 2020-07-20
Publication date: 2020-12-04
Anticipated expiration: 2040-07-20
Also published as: CN112034774B

Abstract

The invention discloses a hot redundancy control method, which adopts a self-adaptive management strategy of a redundant channel and mainly comprises the following steps: the system periodically acquires a fault identification strategy based on the signal state difference and time statistics of the redundant input channels, an input and output channel diagnosis strategy based on the channel incidence relation and diagnosis information of an output channel diagnosis strategy based on output feedback monitoring; and determining the fault by adopting a probabilistic reasoning mode, automatically shielding the work of a fault channel if the fault is an internal fault, realizing the functions of output control and signal feedback by using a mirror image channel, and sending alarm information to a user program if the fault is determined to be an external fault. The invention has the beneficial effects that: when the system fails, the internal automatic switching is realized so as to ensure the normal function of the whole system. The system realizes automatic switching without shutdown. The control method has the advantages of high accuracy of identifying the faults of the input and output channels, high fault positioning speed and timely fault processing.

Description

Hot redundancy control method

Technical Field

The invention relates to the field of thermal redundancy control, in particular to a thermal redundancy control method.

Background

The method of the hot redundancy control system is a key technology for realizing control of a complex system, and in a distributed control system with high requirement on reliability, when the system is required to realize automatic switching without shutdown, redundancy control is often required to be introduced, and internal automatic switching is realized when the system fails, so that the normal function of the whole system is ensured.

For a hot redundancy control system, a dedicated controller supporting hot redundancy control and a hot redundancy control strategy are the core of the control system. The hot redundancy function of the distributed control system comprises basic requirements of communication channel redundancy, input channel redundancy, control output channel redundancy and the like. In various hot redundancy control systems, automatic failure identification and handling of communication channels, control input and output channels is a core key technology. If the controller identifies the input and output channel faults more accurately and more quickly, the more timely the fault is processed, the less the influence on the overall control performance is. Aiming at the hot standby redundancy control requirement with higher reliability requirement, manufacturers at home and abroad respectively put forward different forms of solutions. The existing solution is simple in fault diagnosis and processing strategy, and fails to integrate enough information, so as to realize thorough and accurate positioning and processing of fault information. Such as dynamically changing process information of channel signals, association change information between channels, etc.

Disclosure of Invention

The invention aims to overcome the defects in the prior art and provide a thermal redundancy control method.

The object of the present invention is achieved by the following technical means. A hot redundancy control method adopts a self-adaptive management strategy of a redundant channel, and mainly comprises the following steps: the system periodically acquires a fault identification strategy based on the signal state difference and time statistics of the redundant input channels, an input and output channel diagnosis strategy based on the channel incidence relation and diagnosis information of an output channel diagnosis strategy based on output feedback monitoring; and determining the fault by adopting a probabilistic reasoning mode, automatically shielding the work of a fault channel if the fault is an internal fault, realizing the functions of output control and signal feedback by using a mirror image channel, and sending alarm information to a user program if the fault is determined to be an external fault.

The invention has the beneficial effects that: when the system fails, the internal automatic switching is realized so as to ensure the normal function of the whole system. The system realizes automatic switching without shutdown. The control method has the advantages of high accuracy of identifying the faults of the input and output channels, high fault positioning speed and timely fault processing.

Drawings

FIG. 1 is a fault identification strategy based on redundant input channel signal state differences and time statistics;

FIG. 2 is an input and output channel diagnostic strategy based on channel associations;

FIG. 3 adaptive management policy for redundant channels;

fig. 4 illustrates an adaptive redundant communication network operation mechanism.

Detailed Description

The invention will be described in detail below with reference to the following drawings:

the existing solution is simple in fault diagnosis and processing strategy, and fails to integrate enough information, so as to realize thorough and accurate positioning and processing of fault information. Such as dynamically changing process information of channel signals, association change information between channels, etc.

The invention mainly designs a thermal redundancy control method, which adopts a self-adaptive management strategy of a redundancy channel and mainly comprises the following steps: the system periodically acquires a fault identification strategy based on the signal state difference and time statistics of the redundant input channels, an input and output channel diagnosis strategy based on the channel incidence relation and diagnosis information of an output channel diagnosis strategy based on output feedback monitoring; and determining the fault by adopting a probabilistic reasoning mode, automatically shielding the work of a fault channel if the fault is an internal fault, realizing the functions of output control and signal feedback by using a mirror image channel, and sending alarm information to a user program if the fault is determined to be an external fault. The following diagnostic functions of 4 external channels are automatically implemented in the operational phase:

1) fault identification strategy based on redundant input channel signal state difference and time statistics

2) Input and output channel diagnosis strategy based on channel incidence relation

3) Output channel diagnostic strategy based on output feedback monitoring

4) Adaptive management strategy of redundant channels.

Fig. 1 shows a fault identification strategy based on the signal state difference and time statistics of redundant input channels, which mainly includes the following steps:

1) after the controller hardware is started, the system periodically reads in input channel data obtained from the mirror controller and compares whether the corresponding channel states are consistent.

2) Under the condition that the channel states are consistent, checking whether the state obtained by the current channel in the previous diagnosis process is an undetermined state; and if the state is not in the undetermined state, completing the diagnosis of the current round, setting the diagnosis conclusion to be normal, and setting the counter to be 0. Otherwise, entering the step 3.

3) And adding 1 to the counter, and setting the channel state as 'normal' if the counter reaches a set value.

4) In the first step of diagnosis results, if the channel states are not consistent, judging whether the current channel state is 'pending', and if the current channel state is 'pending', adding 1 to the accumulated difference time; and when the accumulated difference time reaches a set value, setting the channel which is changed recently as normal and setting the other channel as abnormal. Otherwise, the step 5 is carried out.

5) Setting the accumulated difference time to be 1, and setting the current channel state to be 'pending'.

Fig. 2 shows an input and output channel diagnosis strategy based on the channel association relationship, which mainly includes the following steps:

1) and (3) after the system is started, reading the channel association configuration of the system, and then periodically executing the steps 2-5.

2) Calculating whether the channel is changed within a set delta T time (configurable by a user), wherein 0 represents unchanged, and 1 represents changed;

3) checking the action change relation one by one according to a channel configuration table;

4) and if the change relation of the channel configuration is inconsistent with the change relation of the channel configuration, reporting all the associated channels to the self-adaptive management strategy unit.

For example, there are two configurations in the system as shown in the following table:

input lane (lane DX1, lane DX2, lane AX3) output lane (lane D1, lane A2)

Input channels (DX 2, DX3, AX4) output channels (A2, A3)

The variation process state in the Δ T time is:

DX1 variable, DX2 variable, DX3 variable, AX3 variable, lane D1 variable, lane A2 variable, lane A2 variable, lane A3 variable

The system reports DX1, DX2, AX3, D1 and A2 channels.

Fig. 3 shows an adaptive management policy for redundant channels, which mainly includes the following steps: the system periodically acquires the diagnostic information of a fault identification strategy 1 based on the signal state difference of redundant input channels and time statistics, an input and output channel diagnostic strategy 2 based on channel incidence relation and an output channel diagnostic strategy 3 based on output feedback monitoring.

And determining the fault by adopting a probabilistic reasoning mode, automatically shielding the work of a fault channel if the fault is an internal fault, and realizing the functions of output control and signal feedback by using a mirror image channel. And if the external fault is confirmed, sending alarm information to the user program. The probabilistic reasoning mode writes a comprehensive diagnostic program according to the following rules:

1) a channel may be considered "normal" if there is a change in an output channel in the associated channel configuration and the corresponding input channel is in a normal and changing state.

2) An input channel is "failed" if all output channels in a configuration item have not changed, but there is a change in the input channel.

3) A channel is "failed" if in a configuration it changes, but the input channel to which it relates does not, and the other associated output channels are all confirmed as normal in the other configuration.

4) If the conclusion is inconsistent, the conclusion of 'normal' and 'fault' is taken as the basis, and the conclusion of more occurrences is taken as the final conclusion.

For example, in the reported DX1, DX2, AX3, D1 and a2 channels, since the a2 channel is changed, it indicates that the DX1 and DX2 channels are normal, and the AX3 channel and the D1 channel may have faults.

FIG. 4 depicts an adaptive redundant communication network operating mechanism: the controllers (1,2) in redundant configuration in the system are connected with a centralized monitoring system 3 at the upper level through a bus. The communication connections are physically independent. The two controllers are connected by an internal high-speed bus 4. In the working process, each controller is provided with an internal uplink data buffer unit (5,7) and a downlink data buffer unit (6,8), and is also internally provided with a communication buffer unit (9,10) of a mirror image controller.

Each controller physical layer diagnoses a physical link connected with the controller in real time and monitors the communication connection state of the mirror image controller in real time; and when the mirror image data are inconsistent or a certain controller monitors that a communication link of the controller is in failure, performing an emergency transmission mechanism of the bus. When the downlink data received by the two controllers are inconsistent, the system automatically judges the data which changes recently as effective downlink data, and simultaneously receives and transmits the data by using an available normal communication channel.

For example, in fig. 4, if the main controller 1 fails in the downlink channel and the mirror controller 2 fails in the uplink data channel, the main controller 1 obtains the downlink data from the mirror communication buffer unit of the mirror controller 2. At the same time, the mirror controller 2 will send data to the centralized monitoring system via the upstream channel.

It should be understood that equivalent substitutions and changes to the technical solution and the inventive concept of the present invention should be made by those skilled in the art to the protection scope of the appended claims.

Claims

1. A method of controlling thermal redundancy, comprising: the self-adaptive management strategy adopting the redundant channel mainly comprises the following steps: the system periodically acquires a fault identification strategy based on the signal state difference and time statistics of the redundant input channels, an input and output channel diagnosis strategy based on the channel incidence relation and diagnosis information of an output channel diagnosis strategy based on output feedback monitoring; and determining the fault by adopting a probabilistic reasoning mode, automatically shielding the work of a fault channel if the fault is an internal fault, realizing the functions of output control and signal feedback by using a mirror image channel, and sending alarm information to a user program if the fault is determined to be an external fault.

2. The thermal redundancy control method according to claim 1, characterized in that: the fault identification strategy based on the redundant input channel signal state difference and the time statistics mainly comprises the following steps:

1) after the hardware of the controller is started, the system periodically reads in input channel data obtained from the mirror image controller and compares whether the corresponding channel states are consistent;

2) under the condition that the channel states are consistent, checking whether the state obtained by the current channel in the previous diagnosis process is an undetermined state; if the state is not in the undetermined state, completing the diagnosis of the current round, setting a diagnosis conclusion to be normal, setting a counter to be 0, and otherwise, entering the step 3);

3) adding 1 to the counter, and setting the channel state to be normal if the counter reaches a set value;

4) in the diagnosis result of the step 1), if the channel states are not consistent, judging whether the current channel state is 'undetermined', and if the current channel state is 'undetermined', adding 1 to the accumulated difference time; when the accumulated difference time reaches a set value, setting a channel which changes recently as normal and setting the other channel as abnormal, otherwise, turning to the step 5);

3. The thermal redundancy control method according to claim 1, characterized in that: the input and output channel diagnosis strategy based on the channel incidence relation mainly comprises the following steps:

1) after the system is started, reading in the channel association configuration of the system, and then periodically executing the steps 2) to 5);

2) calculating whether the channel is changed within the set delta T time, wherein 0 represents unchanged, and 1 represents changed;

4. The thermal redundancy control method according to claim 1, characterized in that: the probabilistic reasoning mode writes a comprehensive diagnostic program according to the following rules:

1) if there is a change in an output channel in the associated channel configuration, and the corresponding input channel is in a normal and changed state, then the channel is considered "normal";

2) if all output channels in a certain configuration item are unchanged, and an input channel is changed, the input channel is in a fault state;

3) if in a certain configuration item, a certain output channel changes, but the input channel related to the certain output channel does not change, and other associated output channels are all confirmed to be normal in other configuration items, the channel is 'failed';