Disclosure of Invention
In order to achieve the purpose, the invention provides a wireless communication method for dynamic remote fault-tolerant reconstruction of computing resources of a cluster avionic system, and solves the problems that in the prior art, IMA systems on a plurality of aircraft platforms cannot communicate with each other, the computing resources cannot be shared, and reconstruction of the IMA systems cannot be dynamically generated.
The technical scheme adopted by the invention is that the wireless communication method for the dynamic remote fault-tolerant reconstruction of the cluster avionics system computing resources comprises the following steps:
in the first step of the method,
the time required for completing one round of communication among all IMA systems is a communication period T; the basic unit of the communication cycle is a time slot, the length of the time slot can be set according to the actual project requirement, and the time slot is set to be 7.8125 ms; in a communication period, each node sends a message once and receives the message for a plurality of times;
in the second step, the first step is that,
a plurality of continuous time slots form a time frame required by communication of each IMA system, and each time frame can ensure that all system state data, other data required by battles and weapon cooperative data are transmitted by the IMA system; the master control node occupies 0 th of the communication period TA time frame TF0Contains t0A time slot; the 1 st subordinate IMA system occupies the 1 st time frame TF1Contains t1Time slots, and so on, the nth slave IMA system occupies the nth time frame TFNContains tnA time slot,; in the above convention, tiIs a natural number, i is an integer;
step three, performing a first step of cleaning the substrate,
whether the master node or the slave node is in an active state only in an allocated time frame, and can only receive messages sent to the master node or all other IMA systems by any other nodes at other times; the master node at a specified time frame TF0The method can not only broadcast messages to all slave nodes to instruct all IMA systems to report health states, resource states and avionic working modes, but also send messages containing avionic functions to be reconstructed and corresponding avionic working modes to designated slave nodes; the two messages respectively form a sending command message and a remote reconstruction command message, the data formats of the two messages are consistent, and the slave nodes respectively specify a time frame TFjJ is a natural number, and a message is sent to the master control node to report the health state of the master control node and each avionic working mode, wherein the message is called a state data message;
step four, performing a first step of cleaning the substrate,
for the sending command message, the "command identifier" bit is 0, the "receiving node ID" is 0, which means that the master node broadcasts the sending command message to all the slave nodes, and each slave node is required to send status information to the master node, and the "CPM status" area bit is meaningless and may be all 0;
in the fifth step, the step of,
for a remote reconfiguration command message, the "command identification" bit is 1; the 'receiving node ID' is a numerical value within the range of 1-127 and represents the node address of the IMA system selected as a reconstruction object by the main control node in the wireless communication network; the CPM state region bit is used for explaining the reconstruction requirement, and comprises a partition program, an avionic working mode and the number of partition programs which need to be realized by reconstruction, and the partition program is actually state data of the failed CPM before the failure occurs;
step six, performing a first step of treatment,
for the status data message, "sending node ID" represents the slave node IMA address of the currently sent status data, "platform health" is used for explaining whether the IMA system is healthy, 1 represents health, and 0 represents fault, which means that the system needs to be remotely reconstructed; it should be noted that as long as the IMA system solves the failure problem through local reconstruction, the bit is still 1, i.e. it is still recorded as a healthy state, and the IMA system will record the failure event and the solution in the local failure log library; next 16 × N1+4+1 bit, which indicates the partition software resident in the first CPM module, indicated by ID, the working mode of the avionics equipment, the number of partition programs and the health state, 1 indicates that the current CPM is healthy, 0 indicates that the current CPM is faulty, and so on, which indicates that the next CPM module is up to the last CPM of the current IMA system, including the relevant states of all backup CPM modules; the local reconstruction count occupies 2 bits and is used for identifying how many times the current fault node has solved the fault problem of the current IMA system through local reconstruction; the 'available backup CPM number' is used for explaining how many healthy and available backup CPM modules are left in the node, and a judgment basis is provided for remote reconstruction.
The present invention is also characterized in that,
the available time overhead of the remote reconstruction in the step six is calculated as follows:
t for ith slave node to operate in networking of isomorphic cluster avionics systemfaultWhen a fault occurs at any moment, after the fault occurs, the node performs preparation activities such as fault detection, fault message generation, fault message sending to a sending buffer area, message removal from the sending buffer area, channel coding, signal modulation and the like on a local processor for a duration of tau 1;
however, the time of message transmission needs to consider tfaultThe length of the time frame of the distance ti of% T is recorded as tau 9, namely after the fault occurs, the fault node needs to wait for a period of time to report the fault information to the master control node, and the fault time T is consideredfaultThe% T is different before and after the time frame ti arrives, and the values of tau 9 are respectively as follows:
the above formula (1) and formula (2) are represented by a unified mathematical model, as follows:
generally τ 1 is contained within τ 9, and τ 1 < τ 9;
after the time frame of the fault node is brought, the fault node transmits a message, the transmission time and the total amount a of the transmitted informationiIs related to the rate R and is marked as tau 2 ═ aiR; when each node is allocated with a transmission time frame, the data requirement is basically met; transmission time frame t, taking into account tolerance problemsiIs slightly larger than tau 2; the time when the signal is transmitted from the fault node to the master node is recorded as tau 3 ═ Li,0C, wherein Li,0Indicating the distance between the ith aircraft and the master control node, and directly communicating within 350km according to any two aircraft, considering that the light speed is 3 x 105km/s, therefore, the value of τ 3 will be less than 1.85 ms;
for a master control node, after receiving a signal, sending the data to a receiving buffer area through signal demodulation and channel decoding, then waiting for a certain time, moving the data out of the buffer area, performing remote reconstruction decision, generating a remote reconstruction command message, sending the remote reconstruction command message to the sending buffer area, moving the message out of the sending buffer area, performing channel coding and signal modulation, wherein the time consumed in the process is called as remote reconstruction command message preparation time and is marked as tau 4; then, the reconstruction node sends a message to the wireless communication system, the sending time and the total sending information akIs related to the rate R and is marked as tau 5 ═ akThe master control node selects the kth slave node as a reconstruction node, i is not equal to k; master control jointThe transmission time of the signal sent to the reconstruction node k by the point 0 is marked as tau 6, and the tau 6 is approximately equal to tau 3 and is less than or equal to 1.85ms as can be known from the analysis;
after receiving the signals, the reconstruction node k carries out signal demodulation and channel decoding, and then sends the data to a receiving buffer area; after a certain time of waiting, the reconstruction node moves the data out of the buffer area and submits the data to a processor to execute remote reconstruction operation, the process is referred to as remote reconstruction for short, and the duration is recorded as tau 7;
the fault node sends the information to the master node quickly, and the time frame of the master node possibly does not arrive far away, and the interval time is recorded as
Period of master-slave communication
The length of a communication time frame (i.e., the number of slots) of the jth node is represented by a master node when j is 0, and represents a slave node when j is not 0; generally, τ 8 ≧ τ 3+
τ 4;
since the remote reconfiguration decision can only be made by the master control node, the remote reconfiguration cost, i.e. the time overhead, is
Δtremote_reconfig=τ9+τ2+τ8+τ5+τ6+τ7 (4)
Substituting the parameters, the formula (4) is changed into:
from the formula (5), when i is n and
when located in ti time frame, the
interval time τ 9 is 0, τ 8 is τ 3+
τ 4, and the remote reconstruction time overhead is minimal at this time frame
Δtremote_reconfig=τ2+τ3+τ4+τ5+τ6+τ7
=an/R+τ3+τ4+a0/R+τ6+τ7 (6)
When i is 1 and
time, interval time
The remote reconstruction time overhead is then the greatest
As can be seen from equations (6) and (7), the closer the failure time is to the arrival time of the next time frame of the master node, the more urgent the remote reconstruction available time is, and the more relaxed the remote reconstruction available time is.
The invention has the beneficial effects that:
the method can centralize IMA systems on a plurality of airplanes to form a distributed homogeneous cluster avionic system;
all IMA avionics system computing resources within the range of the cluster avionics system can be uniformly monitored and managed;
the method can dynamically select reconstruction nodes and execute remote fault-tolerant reconstruction under the condition that a certain IMA system computing resource fails;
the operational capacity of the cluster of aircraft and the probability of completing a combat mission may be maintained under computing resource failure conditions.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
As shown in fig. 1 to 8, the homogeneous configuration cluster avionics system refers to avionics systems distributed on each aircraft platform and all being IMA-type systems, and the IMA systems communicate in a master-slave mode, wherein an IMA system of a certain aircraft serves as a master node of a wireless communication network of the cluster avionics system, and the rest serve as slave nodes.
The invention discloses a wireless communication method for dynamic remote fault-tolerant reconstruction of cluster avionics system computing resources, which comprises the following steps of:
in the first step of the method,
the time required for completing one round of communication among all IMA systems is a communication period T; the basic unit of the communication cycle is a time slot, the length of the time slot can be set according to the actual project requirement, and the time slot is set to be 7.8125 ms; in a communication period, each node sends a message once and receives the message for a plurality of times;
in the second step, the first step is that,
a plurality of continuous time slots form a time frame required by communication of each IMA system, and each time frame can ensure that all system state data, other data required by battles and weapon cooperative data are transmitted by the IMA system; the main control node occupies the 0 th time frame TF of the communication period T0Contains t0A time slot; the 1 st subordinate IMA system occupies the 1 st time frame TF1Contains t1Time slots, and so on, the nth slave IMA system occupies the nth time frame TFNContains tnA time slot,; in the above convention, tiIs a natural number, i is an integer;
step three, performing a first step of cleaning the substrate,
whether the master node or the slave node is in an active state only in an allocated time frame, and can only receive messages sent to the master node or all other IMA systems by any other nodes at other times; the master node at a specified time frame TF0The method can not only broadcast messages to all slave nodes to instruct all IMA systems to report health states, resource states and avionic working modes, but also send messages containing avionic functions to be reconstructed and corresponding avionic working modes to designated slave nodes; the two messages respectively form a sending command message and a remote reconstruction command message, the data formats of the two messages are consistent, and the slave nodes respectively specify a time frame TFjJ is a natural number, and a message is sent to the master control node to report the health state of the master control node and each avionic working mode, wherein the message is called a state data message;
step four, performing a first step of cleaning the substrate,
for the sending command message, the "command identifier" bit is 0, the "receiving node ID" is 0, which means that the master node broadcasts the sending command message to all the slave nodes, and each slave node is required to send status information to the master node, and the "CPM status" area bit is meaningless and may be all 0;
in the fifth step, the step of,
for a remote reconfiguration command message, the "command identification" bit is 1; the 'receiving node ID' is a numerical value within the range of 1-127 and represents the node address of the IMA system selected as a reconstruction object by the main control node in the wireless communication network; the CPM state region bit is used for explaining the reconstruction requirement, and comprises a partition program, an avionic working mode and the number of partition programs which need to be realized by reconstruction, and the partition program is actually state data of the failed CPM before the failure occurs;
step six, performing a first step of treatment,
for the status data message, "sending node ID" represents the slave node IMA address of the currently sent status data, "platform health" is used for explaining whether the IMA system is healthy, 1 represents health, and 0 represents fault, which means that the system needs to be remotely reconstructed; it should be noted that as long as the IMA system solves the failure problem through local reconstruction, the bit is still 1, i.e. it is still recorded as a healthy state, and the IMA system will record the failure event and the solution in the local failure log library; next 16 × N1+4+1 bit, which indicates the partition software resident in the first CPM module, indicated by ID, the working mode of the avionics equipment, the number of partition programs and the health state, 1 indicates that the current CPM is healthy, 0 indicates that the current CPM is faulty, and so on, which indicates that the next CPM module is up to the last CPM of the current IMA system, including the relevant states of all backup CPM modules; the local reconstruction count occupies 2 bits and is used for identifying how many times the current fault node has solved the fault problem of the current IMA system through local reconstruction; the 'available backup CPM number' is used for explaining how many healthy and available backup CPM modules are left in the node, and a judgment basis is provided for remote reconstruction.
The available time overhead of the remote reconstruction in the step six is calculated as follows:
t for ith slave node to operate in networking of isomorphic cluster avionics systemfaultWhen a fault occurs at any moment, after the fault occurs, the node performs preparation activities such as fault detection, fault message generation, fault message sending to a sending buffer area, message removal from the sending buffer area, channel coding, signal modulation and the like on a local processor for a duration of tau 1;
however, the time of message transmission needs to consider tfault% T distance ti time frameThe length of (c) is recorded as τ 9, that is, after a fault occurs, the fault node needs to wait for a period of time before reporting the fault information to the master node, and the fault time t is consideredfaultThe% T is different before and after the time frame ti arrives, and the values of tau 9 are respectively as follows:
the above formula (1) and formula (2) are represented by a unified mathematical model, as follows:
generally τ 1 is contained within τ 9, and τ 1 < τ 9;
after the time frame of the fault node is brought, the fault node transmits a message, the transmission time and the total amount a of the transmitted informationiIs related to the rate R and is marked as tau 2 ═ aiR; when each node is allocated with a transmission time frame, the data requirement is basically met; transmission time frame t, taking into account tolerance problemsiIs slightly larger than tau 2; the time when the signal is transmitted from the fault node to the master node is recorded as tau 3 ═ Li,0C, wherein Li,0Indicating the distance between the ith aircraft and the master control node, and directly communicating within 350km according to any two aircraft, considering that the light speed is 3 x 105km/s, therefore, the value of τ 3 will be less than 1.85 ms;
for a main control node, after receiving a signal, the main control node transmits the data to a receiving buffer area through signal demodulation and channel decoding, then after waiting for a certain time, the data moves out of the buffer area to perform remote reconstruction decision, a remote reconstruction command message is generated, the remote reconstruction command message is transmitted to a transmitting buffer area, the message moves out of the transmitting buffer area to perform channel coding and signal modulation, and the time consumed in the process is called as the remote reconstruction command messageThe text preparation time is recorded as tau 4; then, the reconstruction node sends a message to the wireless communication system, the sending time and the total sending information akIs related to the rate R and is marked as tau 5 ═ akThe master control node selects the kth slave node as a reconstruction node, i is not equal to k; the transmission time of the signal sent to the reconstruction node k by the main control node 0 is recorded as tau 6, and the tau 6 is approximately equal to tau 3 and is less than or equal to 1.85ms as can be known from the analysis;
after receiving the signals, the reconstruction node k carries out signal demodulation and channel decoding, and then sends the data to a receiving buffer area; after a certain time of waiting, the reconstruction node moves the data out of the buffer area and submits the data to a processor to execute remote reconstruction operation, the process is referred to as remote reconstruction for short, and the duration is recorded as tau 7;
the fault node sends the information to the master node quickly, and the time frame of the master node possibly does not arrive far away, and the interval time is recorded as
Period of master-slave communication
The length of a communication time frame (i.e., the number of slots) of the jth node is represented by a master node when j is 0, and represents a slave node when j is not 0; generally, τ 8 ≧ τ 3+
τ 4;
since the remote reconfiguration decision can only be made by the master control node, the remote reconfiguration cost, i.e. the time overhead, is
Δtremote_reconfig=τ9+τ2+τ8+τ5+τ6+τ7 (4)
Substituting the parameters, the formula (4) is changed into:
from the formula (5), when i is n and
when the time interval is located in the ti time frame, the
interval time τ 9 is 0, and τ 8 is τ 3+
τ 4, and the remote process is repeatedThe overhead of the construction time is minimum, that is
Δtremote_reconfig=τ2+τ3+τ4+τ5+τ6+τ7
=an/R+τ3+τ4+a0/R+τ6+τ7 (6)
When i is 1 and
time, interval time
The remote reconstruction time overhead is then the greatest
As can be seen from equations (6) and (7), the closer the failure time is to the arrival time of the next time frame of the master node, the more urgent the remote reconstruction available time is, and the more relaxed the remote reconstruction available time is.
The invention inherits the concept of ASAAC IMA avionics design, and can support remote communication by slightly modifying a single machine IMA system, thereby adapting to the requirements of IMA avionics system communication networking on a plurality of airplane platforms; by adding a reconfiguration management function, the interaction behavior between reconfiguration management components is designed, and local and remote fault-tolerant reconfiguration is supported; determining a communication mode among IMA systems in the same configuration cluster avionics system, establishing a performance model of remote fault-tolerant reconstruction, and performing performance analysis of remote reconstruction.
The existing IMA system is generally composed of a plurality of general processing modules CPM and a mass storage module MMM, and each module is resident with three-layer stack IMA software, namely ① module support layer software which provides board-level support programs such as a driver of a hardware module, ② operating system layer software which provides basic software such as an operating system, and ③ application layer software which provides avionic function partition programs.
The IMA system is reformed, and the specific reformation comprises three aspects:
the method comprises the steps of adding middleware on an operating system layer, as shown in fig. 1, wherein the middleware comprises ① basic middleware which is responsible for packaging communication and concurrency mechanisms of a partitioned real-time operating system (PRTOS) and creating a reusable network programming component, ② distributed middleware which is responsible for managing each IMA system resource, multiplexing an application programming interface and the network programming component, automatically executing and expanding network programming capacity of the partitioned real-time operating system (PRTOS) and allowing writing of a distributed application program, and ③ universal middleware service which is responsible for managing distributed tasks such as log, event notification, real-time scheduling, concurrency control and the like of various resources in a distributed homogeneous cluster avionics system.
Secondly, a reconfiguration management component is added in the operating system layer, as shown in fig. 1, and is responsible for reconfiguration management in a corresponding authority range.
The homogeneous cluster avionics system works in a master-slave mode, namely one node is used as a master node, the other nodes are used as slave nodes, communication is carried out according to a fixed time slot distribution mode, and the homogeneous cluster avionics system is used as an IMA system of the slave nodes, and only a local runtime blueprint RTBP database, a cluster software database and a fault log database reside in a mass memory of the homogeneous cluster avionics system to support local fault-tolerant reconstruction; in the IMA system serving as the master control node, a resource state database and an avionic working mode database reside in a mass storage of the IMA system, and remote fault-tolerant reconstruction is supported, as shown in fig. 2.
The fault-tolerant reconfiguration management of the same-configuration cluster avionics system is responsible for resource management, working mode control, working parameter setting, working state monitoring, system information scheduling, system redundancy and reconfiguration management, alarm processing and the like of all IMA systems in a cluster range.
The invention establishes a hierarchical reconstruction management architecture which is divided into 3 levels from bottom to top, namely resource level reconstruction management RL-GSM, platform level reconstruction management PL-GSM and cluster level reconstruction management CL-GSM. The resource level reconfiguration management RL-GSM example is responsible for monitoring the resource utilization state and the health state on the CPM or MMM module, and managing the resources on the CPM or MMM. The platform level reconfiguration management PL-GSM instance is then responsible for managing, monitoring and controlling all resources and their status of the IMA system in which it is located. As for the CL-GSM example of cluster-level reconfiguration management, all resources and states of the whole cluster avionics system are monitored, controlled and managed, and the CL-GSM example is a decision maker and manager for remote dynamic fault-tolerant reconfiguration of the cluster avionics system. One cluster-level reconfiguration management CL-GSM instance is responsible for managing a plurality of platform-level management PL-GSM, and one platform-level management PL-GSM in turn manages a plurality of resource-level reconfiguration management RL-GSM. In the invention, each level of reconstruction management instance is resident in an operating system layer of IMA software; each CPM and MMM of all IMA systems reside with a resource level reconfiguration management RL-GSM example; the platform level reconfiguration management PL-GSM example can only run on MMM, and each IMA system has one and only one; the cluster level reconfiguration CL-GSM instances have one and only one in the overall cluster avionics system, residing on the MMM hosting the IMA system. These reconfiguration management instances form the management and control logic shown in FIG. 5 based on resource membership.
For each level of reconfiguration management, the reconfiguration management must complete 3 aspects of work tasks, namely, health monitoring HM, fault management FM and configuration management CM, and the three work tasks are respectively completed by independent software components. The health monitoring HM is used for monitoring the use condition of avionic resources, evaluating the health state of the IMA system and the multi-platform avionic system and maintaining a fault log library. The fault management FM includes various fault handling measures, such as positioning, isolating, limiting, and the like, and can prevent a fault from occurring or limit a fault diffusion range, and ensure that the IMA system after the fault can continuously run for a period of time, so as to provide a time for completing processing operations such as reconstruction. The configuration management CM realizes the establishment of system initial configuration and the management and control of the system reconfiguration process. The health monitoring HM, the fault management FM and the configuration management CM may interact with each other in the same hierarchy, may interact with each other in different hierarchies, and may even interact with a database to obtain necessary data, as shown in fig. 6.
When a CPM module of an IMA system in the range of the cluster avionics system fails, resource level health monitoring on the module, namely a health monitoring component RL-GSM-HM in resource level reconfiguration management, performs fault filtering, association and confirmation according to the current configuration and fault filtering algorithm of the system. The RL-GSM-HM will then inform this level of fault management about this fault and also about the diagnostic data. The fault targeted by the invention is permanent rather than instantaneous and can not be repaired by restarting, so the resource level fault management RL-GSM-FM reports the fault to the upper level health monitoring PL-GSM-HM. The PL-GSM-HM informs the fault of the current-level fault management PL-GSM-FM, and inquires a local RTBP database by the PL-GSM-HM to determine fault processing behaviors, such as fault covering, fault isolation and the like. If there is a feasible fault handling behavior for the fault, i.e. the fault can be resolved at this level, the PL-GSM-FM sends a change configuration request to the peer configuration management PL-GSM-CM.
And after the platform level configuration management PL-GSM-CM receives the configuration change request, sending a current configuration stop command to the RL-GSM-CM where the fault CPM resides. And the RL-GSM-CM is responsible for stopping all partition programs on the fault CPM, removing the association relation between the partition programs and the communication channel, closing the communication channel from the fault module to other related CPM/MMM modules, destroying all partition programs on the fault CPM, feeding back the status message of stopped configuration after the configuration is stopped to the RL-GSM-CM, and simultaneously sending a configuration stop command message to the locally resident RL-GSM-HM and RL-GSM-FM to stop the current configuration of the RL-GSM-CM and the RL-GSM-FM so as to reset the RL-GSM-CM to the initial configuration state.
For the PL-GSM-FM selected backup CPM, which works in the initial state, namely warm backup state, there is no resident OS, middleware and any partition program except the resident module support layer program and resource level reconfiguration hypervisor, so there is no need to stop the current configuration. The PL-GSM-CM first issues a load new configuration command message to the RL-GSM-CM. After receiving the command, the configuration management RL-GSM-CM requests and controls the backup CPM to download the OS, the middleware and all the partition programs originally running on the fault CPM from the cluster software database, establishes a communication channel from the backup module to other related modules, associates the partition programs with the communication channel, and returns a message that the new configuration of the PL-GSM-CM is loaded after the work is finished. The PL-GSM-CM receives the message and then sends a new configuration operation command to the RL-GSM-CM. After the new configuration is run on the backup CPM, the RL-GSM-CM replies with a PL-GSM-CM new configuration running message to complete the local reconfiguration process within the scope of the IMA system.
When the fault can not be solved by executing local reconstruction in the range of the IMA system of the fault, the PL-GSM-FM further reports the fault to cluster level management so as to carry out dynamic fault-tolerant reconstruction in the range of the cluster avionics system. After receiving and confirming the fault, the cluster level health monitoring (denoted as CL-GSM-HM) submits the fault information to the current level fault management CL-GSM-FM. And the cluster level fault management carries out real-time dynamic reconfiguration decision according to the health state and the resource utilization state of other IMA systems, determines a new system configuration, sends the new system configuration to the level configuration management, and then executes reconfiguration.
In summary, when local fault-tolerant reconfiguration is performed on the cluster avionics system, static configurations defined in advance by each IMA system are relied on, and the configurations are stored in a local RTBP database and can be consulted as required, so that system reconfiguration is supported. Once remote fault-tolerant reconstruction is required, a fault-tolerant reconstruction decision in the static configuration search form cannot be realized because a configuration defined in advance does not exist. Considering that the health state and the resource utilization state of each IMA system of the cluster avionics system are dynamically changed, the CPM module can be changed from the health state to the fault state, and the backup CPM module can be changed from the idle state to the normal working state, so that a dynamic reconfiguration decision needs to be made according to the resource state in the cluster avionics system.
The method comprises the steps that whether a backup CPM module exists in a target IMA system of ①, whether the backup CPM module of ② is healthy or not and whether avionic software is distributed to a backup CPM module of ③ or not, namely, the backup CPM module is in a normal working state or not, wherein the factors influencing the remote reconfiguration time mainly comprise the time intervals of a main control node, a fault node and a reconfiguration node, in order to enable the avionic system of the cluster of the same configuration to recover and execute the influenced avionic function as soon as possible so as to complete a combat mission, the fault node needs to receive data sent back by the migrated avionic function as soon as possible.
In addition, each IMA system has avionics applications that affect flight safety, such as flight control, cockpit display control, aircraft management, and the like. Considering that the safety of reconstruction is more important than the reliability of tasks, when the CPM where the avionics applications are located fails, only local degradation reconstruction can be performed, but remote reconstruction cannot be performed.
By combining the above factors, the invention designs a dynamic fault-tolerant reconstruction decision algorithm as shown in fig. 7, which is specifically performed according to the following steps:
1) a fault processing function module PL-GSM-FM in a reconstruction management component on a fault airplane receives fault information sent by a health monitoring function module PL-GSM-HM in a reconstruction management component of the level;
2) the fault processing function module PL-GSM-FM judges whether the IMA system at the current level has a healthy, unused and directly-affiliated backup CPM module, if so, the step 3) is executed, and if not, the step 4) is executed;
3) directly carrying out local reconstruction on a first IMA system directly-affiliated CPM module which meets the conditions on the fault aircraft, and carrying out step 16) after the local reconstruction is finished;
4) judging whether any safety-critical avionics application programs such as flight control, airplane management, cockpit display control and the like are allowed on a faulted CPM module, if yes, executing 5), and if not, executing 6);
5) reselecting a CPM module residing with non-safety-critical avionics applications in an IMA system of a fault aircraft, executing local degradation reconstruction, and executing step 16) after local degradation reconstruction is completed;
6) the PL-GSM-FM of the fault airplane reports the fault information to a health monitoring function module CL-GSM-HM of a cluster-level reconstruction management component of the master control airplane;
7) the CL-GSM-HM carries out fault filtering and confirmation and sends fault information to a fault processing function module CL-GSM-FM in the level reconstruction management component;
8) the cluster-level CL-GSM-FM starts to sequentially traverse each IMA system according to the designed communication time slot sequence;
9) the cluster-level CL-GSM-FM judges whether the IMA system traversed currently is a fault IMA system, if so, the fault IMA system enters 10), and if not, the fault IMA system enters 11);
10) judging whether all the slave IMA systems on the non-master aircraft have been traversed, if so, executing 12), and otherwise, entering 11);
11) if the currently traversed non-fault IMA system is healthy and unused, directly backing up the CPM module, if yes, executing 14), and if not, executing 15);
12) performing step 5);
13) performing step 8);
14) performing remote reconstruction on the first IMA system directly-affiliated CPM module meeting the conditions, and executing the step 16) after the remote reconstruction is completed;
15) execution 10);
16) and finishing the reconstruction decision.
The above description is only for the preferred embodiment of the present invention, and is not intended to limit the scope of the present invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention shall fall within the protection scope of the present invention.