Keywords

1 Introduction

The current domestic power industry substation training simulation system mainly uses Virtools, VRML and other traditional simulation tools to construct one-person–operation desktop application [1], which is poor to realism, presence, expression and interaction and not applicable to multi-user collaboration [2]. With the rapid development of virtual reality technology, the substation collaborative virtual simulation training based on the distributed network came into being. In recent years, the problem of collaborative simulation in distributed virtual reality network has been paid close attention by researchers [3, 4]. However, there is limited research on substation operation and maintenance simulation system supporting multi-user cooperative interaction in distributed network environment [5].

This paper proposes a process modeling technology based on hybrid heterogeneous hierarchical modeling technology [6] and a concurrent control model based on Petri net shared subnet synthesis technology [7]. The hierarchical process modeling of power maintenance tasks, distributed collaborative concurrency control, asynchronous message delay strategy and other key technologies are discussed. At last, a substation collaborative simulation training system was developed.

2 Collaborative Simulation Training System Model for Substation

The substation multi-user collaborative simulation training system in accordance with the Computer-Supported Cooperative Work [8] design ideas can be divided into three levels. The principal-factors model is shown in Fig. 1.

Fig. 1.
figure 1

Substation collaborative simulation training system model

Full size image

The collaborative interaction layer is located at the top, providing the user with an interactive substation virtual environment. The presentation driven layer is the virtual reality terminal function core which is driven by the professional virtual reality 3D engine. It is deployed in each user’s client machine, responsible for driving the virtual reality substation, dynamic loading and rendering the substation scene and equipment resources. The service control layer is responsible for managing the task scenario of the substation simulation system, carrying out concurrency control for the collaborative users in the distributed environment and electrical simulation calculation.

3 Key Techniques to the System Implementation

The collaborative training system should be based on the object-oriented continuous-discrete event hybrid system modeling and simulation method. And the following ideas of collaborative virtual environment technology are used to solve the problem of object modeling, behavior concurrency control, data consistency and synchronization problem: hybrid heterogeneous hierarchical process modeling, distributed cooperative concurrency control and distributed network synchronous delay control.

3.1 Hybrid Heterogeneous Hierarchical Process Modeling of Power Operation

In this paper, the hybrid heterogeneous hierarchical modeling is introduced into the modeling process of the substation simulation process model. The virtual simulation operation process model of the substation is abstracted into three levels [9]: task layer model, job layer model and equipment layer model to construct task-based hierarchical power simulation training process model framework.

Task Layer Modeling Method

Power operation procedure comprises certain power electrical devices operating state conversion and operation changes. The task layer model uses a series of operational logic functions (AND, OR and etc.) to describe the sequential control logic under the constraints of the process control. Depending on whether the operating status and operability of the equipment is available, the functions are set to “1” or “0”.

In this paper, we take replacing the cable task operation process for modeling examples of the process of building the hybrid heterogeneous hierarchical levels of the method. As shown in Fig. 2, the equivalent electrical wiring diagram of the main electrical equipment of the cable replacement task is described.

Fig. 2.
figure 2

Cable replacement operation equivalent electrical diagram

Full size image

The initial state of the main electrical equipment before operation is: A = 1, C = 1, E = 0. After the first operation is completed, the status of each device is A = 0, C = 0, E = 1, B = 1, D = 1, and O = 0.

Job Layer Modeling Method

The job layer model describes the operational behavior and execution flow of the main objective of the main electrical equipment. This paper builds the job model based on the Petri net.

Definition 1.

Let \( PN = \left( {P,T,F,M_{0} } \right) \) be an ordinary Petri net system, where P denotes the places, T denotes the transition, F denotes the flow relation, and M0 denotes the initial state of Petri net.

Definition 2.

Let \( M_{1} = CU + M_{0} \) be the relationship between the initial and reachable state in the Petri net [10, 11], where M0 denotes the initial identity of the Petri net, M1 denotes the reachability of the Petri net, C denotes the correlation matrix of the Petri net, U denotes the transition sequence vector.

According to the Definition 2, the job operation starts from the initial condition M0, and the transition is judged by the value of the U transition sequence, that is, the operation of the behavior of the event. If the transition has the right to occur under the current identification condition, the corresponding operation has the permission, otherwise the change is not allowed to occur. Specific steps are shown in Fig. 3.

Fig. 3.
figure 3

Job layer state transition algorithm

Full size image

In this paper, we take the operation of the task 1 of Fig. 2: “close 3# pole top interconnection switch isolation switch and circuit breaker for switching operation” as an example of job layer modeling example. The corresponding Petri net is shown in Fig. 4.

Fig. 4.
figure 4

Petri nets of the job layer model

Full size image

Equipment Layer Modeling Method

Equipment atomic action refers to the operational response of the movable parts of the equipment to the operation according to the equipment control system and the mechanical lock. And the action behavior of the equipment is a set of limited states. Thus, the atomic action layer model of the equipment can be represented by a finite state machine.

3.2 Concurrency Control Model Based on Petri Shared Subnets

As shown in Fig. 4, all of the substation simulation operations can be modeled by Petri nets, even if the multi-user collaboration scenario is no exception.

Definition 3.

Let \( {\text{PN}}_{\text{i}} = \left( {{\text{P}}_{\text{i}} ,{\text{T}}_{\text{i}} ,{\text{F}}_{\text{i}} ,{\text{M}}_{{0{\text{i}}}} } \right)\;\;\left( {{\text{i}} = 1,2} \right) \) are two Petri net systems that meet the following conditions:

  • \( P = P_{1} \cup P_{2} ,\;P_{1} \cap P_{2} \ne \emptyset , \)

  • \( T = T_{1} \cup T_{2} ,\;T_{1} \cap T_{2} \ne \emptyset , \)

  • \( F = F_{1} \cup F_{2} \). Then PN1 and PN2 are the shared subnets of PN.

It can be seen from the Definition 3 that in the virtual substation collaborative simulation operation, the multi-user cooperates with each other to perform a subnet, and the multiple subnets together constitute the whole operation. At the same time, the individual operation follows the state transition algorithm of the synthetic Petri net.

Figures 5 and 6 show the two shared subnets of the Petri net model in Fig. 4. PN1 and PN2 can represent the job models of the two service personnel in the same collaboration scenario. The two maintenance personnel perform a collaboration to complete the PN network.

Fig. 5.
figure 5

Petri net shared subnet PN1

Full size image
Fig. 6.
figure 6

Petri net shared subnet PN2

Full size image

It can effectively achieve the concurrency control in the distributed virtual reality environment by maintaining a copy of the synthetic Petri net in the distributed terminal client and being consistent with the center Petri net in the server. Besides, it greatly reduces the synchronization delay caused by the network condition. The specific strategy is shown in Fig. 7.

Fig. 7.
figure 7

Asynchronous message concurrency control model

Full size image

Both the local client and the central server maintain a unified set of synthetic Petri network state data. All users first access the local Petri network for operation judgement. If the condition is enabled, the user will synchronize the execution and send the transition status message to the central server for updates. The interactive data such as the behavior, the language, and the input of the operator will be sent in the form of an asynchronous message later.

4 System Implementation and Evaluation

According to the virtual reality substation collaborative simulation training system model, the typical application of simulation training scene for UHV substation is realized. Taking the substation switching operation training as an example, the user is located in the corresponding virtual scene and operates according to the operation procedure of the switching operation.

In the actual evaluation process, six tests were carried out at different points in the day. As shown in Fig. 8, the delay between the local response time of the system and the update time in the local area network environment were kept in 0.01 s, while the delay between the local response time and the update response time in a heterogeneous network environment is extended to about 0.03 s. According to the current average refresh rate of VR terminal equipment which is about 90 fps, there is a certain frame loss in the distributed network environment. But it is still within the acceptable range.

Fig. 8.
figure 8

Time delay comparison chart

Full size image

5 Summary

Regarding the imperfection and poor collaborative interaction ability in existing substation simulation training system, this paper proposes a distributed virtual reality simulation training system model for multiuser cooperation in substation maintenance which adopts virtual reality, cooperative interaction and other technologies. A process modeling technology based on hybrid heterogeneous hierarchical modeling technology and a concurrency control model based on Petri net shared subnet synthesis technology is discussed. This paper also introduces the key technologies such as hierarchical process modeling, distributed collaborative concurrency control and asynchronous message delay control strategy. At last, the substation collaborative simulation training system was developed by using Unity3D virtual reality engine technology, which realized the typical scene application of multi-user coordination operation of power overhaul operation. This result has solved the training dilemma of substation multi - type participation in maintenance simulation training, and has some reference significance for the development of other multi-user cooperation in electric power operation.