A collaborative control method and system for negative pressure systems based on modular assembly

Abstract

This invention relates to the field of negative pressure control technology, and proposes a collaborative control method and system for negative pressure systems based on modular assembly. The method includes: adjacent nodes in a negative pressure pipeline network engaging in handshake communication, and determining the influence coefficient of each node based on the flow relationship between the nodes after communication; predicting the pressure fluctuation curves of neighboring nodes based on the action intentions of all neighboring nodes around the current node, and weighting and superimposing the influence coefficients according to the pressure fluctuation curves to obtain the feedforward compensation signal of the current node; performing virtual inertia before the feedforward compensation signal is superimposed on the current node; and compensating the current node according to the corrected collaborative control quantity. This invention can improve the efficiency of collaborative control of a negative pressure system based on modular assembly.

Description

Technical Field

[0001] This invention relates to the field of negative pressure control technology, and in particular to a method and system for coordinated control of negative pressure systems based on modular assembly. Background Technology

[0002] In modularly assembled negative pressure pipeline systems, functional nodes such as suction sources and terminal gas consumption points are flexibly combined and deployed according to process requirements. These nodes are coupled together through the pipeline topology, forming a complex fluid dynamic network. Traditional negative pressure control methods generally employ a single-node independent adjustment strategy, where each node performs closed-loop control solely based on the feedback deviation of its local pressure sensor, lacking information exchange and coordination mechanisms. However, due to the significant time lag and coupling inherent in the transmission of pressure fluctuations within the pipeline network, any adjustment to the operating conditions of any node will propagate along the pipeline as a pressure wave, causing pressure disturbances in adjacent branches. Each node, upon sensing the disturbance, performs its own compensation action, which, due to misaligned response timing, further creates a chain reaction of oscillations, resulting in the overall system pressure remaining in a fluctuating state for extended periods, making it difficult to converge to a stable operating condition. Especially in industrial scenarios with frequent operating condition switching or sudden load changes, this fragmented control not only significantly increases system energy consumption but also accelerates the mechanical wear of end-effectors such as fans and valves, reducing system lifespan.

[0003] To address the aforementioned coupling interference problem, some studies have attempted to introduce feedforward compensation strategies, but existing solutions still have significant limitations. Traditional feedforward methods mostly rely on offline-calibrated fixed pipeline models or transfer functions, making it difficult to adaptively match the actual operating conditions of frequently changing pipeline topologies in modular assembly scenarios. They fail after system expansion or reconstruction. Existing technologies lack an active prediction mechanism for the intentions of neighboring nodes, only triggering compensation responses after disturbances occur and are captured by sensors, leaving the control lag problem unresolved. Conventional compensation signals are directly superimposed onto the local control loop, failing to consider the different requirements for compensation response speeds due to the functional roles of different nodes. This can easily introduce additional pressure spikes due to compensation overshoot, or cause momentary insufficient suction capacity due to slow response. Therefore, improving the efficiency of collaborative control of a modularly assembled negative pressure system has become an urgent problem to be solved. Summary of the Invention

[0004] This invention provides a method and system for coordinated control of negative pressure systems based on modular assembly, in order to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides a collaborative control method for a negative pressure system based on modular assembly, comprising: S1. Adjacent nodes in the negative pressure pipeline network communicate by handshaking, and determine the influence coefficient of each node based on the process relationship between the nodes after communication. S2. Based on the action intentions of all neighboring nodes around this node, predict the pressure fluctuation curves of the neighboring nodes, and weight and superimpose the influence coefficients according to the pressure fluctuation curves to obtain the feedforward compensation signal of this node. S3. Before the feedforward compensation signal is superimposed onto the local node, virtual inertia is calculated. S4. Compensate the current node according to the corrected cooperative control quantity.

[0006] In a preferred embodiment, adjacent nodes in the negative pressure pipeline network perform handshake communication, and determine the influence coefficient of each node based on the process relationship between the nodes after communication, including: Adjacent nodes exchange their unique identifiers and functional role labels that represent their role in the process flow, wherein the functional role label identifies at least a suction source or a terminal gas consumption point. Based on the neighbor functional role tags obtained from the exchange, query the pre-established process flow ontology knowledge graph, and deduce the upstream and downstream logical relationships between this node and neighboring nodes. Based on the upstream and downstream logical relationship, an influence weight coefficient is calculated for each neighbor node.

[0007] In a preferred embodiment, calculating an influence weight coefficient for each neighbor node based on the upstream and downstream logical relationship includes: Based on the upstream and downstream logical relationship, neighboring nodes are classified into upstream and downstream nodes on the same branch, cross-branch related nodes, or independent branch nodes, and the influence weights of the corresponding levels are assigned to neighboring nodes of different categories.

[0008] In a preferred embodiment, adjacent nodes in the negative pressure pipeline network perform handshake communication, and determine the influence coefficient of each node based on the process relationship between the nodes after communication, further including: During the handshake communication phase, the network access location description information reported by neighboring nodes is received; By combining the functional role label of this node with the pipeline access location description information of the neighboring node, it is determined whether there is a direct process material flow relationship between the two nodes. Based on the determination that there is a direct process material flow relationship, the neighboring nodes from which process materials flow out are marked as upstream influence sources, and the neighboring nodes from which process materials flow in are marked as downstream influence objects. The upstream influence sources are assigned an influence coefficient that is higher than that of the downstream influence objects.

[0009] In a preferred embodiment, the influence coefficient is calculated using the following formula: ; In the formula, For nodes Neighboring nodes The influence coefficient of the allocation Neighboring nodes When nodes are classified as upstream or downstream nodes of the same branch, the baseline value is used. For nodes with neighboring nodes Topological distance, For nodes with neighboring nodes Topological distance, For nodes The set of all neighboring nodes, is the ordinal number of the neighboring node. It is a natural constant. Neighboring nodes When nodes are classified as upstream or downstream nodes of the same branch, the baseline value is used. This is a preset attenuation factor based on the physical characteristics of the pipeline network.

[0010] In a preferred embodiment, predicting the pressure fluctuation curves of neighboring nodes based on the action intentions of all neighboring nodes surrounding the current node includes: This node receives the action intent instructions pre-broadcast by each neighboring node before executing the control action, and extracts the expected value of the target frequency to be adjusted by the neighboring node and the action execution timestamp from the action intent instructions; Based on the expected target frequency, determine the wind turbine speed change range of neighboring nodes, and align the actions of neighboring nodes in time sequence by combining the action execution timestamps. For each neighboring node, based on the fan speed change range, the corresponding pipeline pressure response characteristic curve of the neighboring node is called to generate the pressure fluctuation curve that the neighboring node is expected to cause at the outlet of this node.

[0011] In a preferred embodiment, the step of weighting and superimposing the influence coefficients according to the pressure fluctuation curve to obtain the feedforward compensation signal of the current node includes: This node obtains the pressure fluctuation curves and influence coefficients of each neighboring node to establish a mapping table between pressure fluctuation curves and influence coefficients. The type of disturbance caused by the actions of neighboring nodes to the negative pressure of this node is determined based on the direction of the pressure fluctuation curve. The disturbance type is classified as positive pressure impact or negative pressure suction. After expanding the pressure fluctuation curve along the time axis, at each time sampling point, the pressure fluctuation value of each neighboring node at the current time is multiplied by its corresponding influence coefficient to obtain the weighted fluctuation component. Then, the weighted fluctuation components of all neighboring nodes are accumulated according to the disturbance type to obtain the feedforward compensation signal.

[0012] In a preferred embodiment, the step of performing virtual inertia calculation before the feedforward compensation signal is superimposed onto the local node includes: This node obtains its own functional role label determined in the handshake communication, and determines the target virtual inertia characteristic to be adopted by this node based on the correspondence between the functional role label and the preset virtual inertia characteristic. The feedforward compensation signal is input to a signal buffer with the target virtual inertia characteristics, and the signal buffer performs time-level flattening on the abrupt amplitude of the feedforward compensation signal. The signal buffer outputs a flattened feedforward compensation signal, which is used as the compensation signal to be superimposed on this node.

[0013] In a preferred embodiment, the signal buffer stage outputs a flattened feedforward compensation signal as the signal to be superimposed onto the compensation signal of this node, and further includes: This node monitors the rate of change of the feedforward compensation signal in real time. When the rate of change exceeds the preset stable change threshold, it determines that there is a sudden change component in the feedforward compensation signal. Extract the magnitude of the mutation component and redistribute the magnitude of the mutation according to the time axis to generate a smooth compensation transition curve with the change slope first increasing and then decreasing. The abrupt component in the original feedforward compensation signal is replaced by the smooth compensation transition curve to form a gradually changing feedforward compensation signal, allowing the node to have a buffer time before performing the compensation action.

[0014] To address the aforementioned problems, the present invention also provides a modularly assembled negative pressure system collaborative control system, the system comprising: The topology handshake and weight calculation module is used for adjacent nodes in a negative pressure pipeline network to perform handshake communication and determine the influence coefficient of each node based on the process relationship between nodes after communication. The intention prediction and feedforward synthesis module is used to predict the pressure fluctuation curves of neighboring nodes based on the action intentions of all neighboring nodes around the current node, and to weight and superimpose the influence coefficients based on the pressure fluctuation curves to obtain the feedforward compensation signal of the current node. The virtual inertia buffer conditioning module is used to perform virtual inertia before the feedforward compensation signal is superimposed onto the local node. The collaborative compensation execution module is used to compensate the local node based on the corrected collaborative control quantity. Compared with the prior art, the present invention has the following beneficial effects: 1. This invention utilizes inter-node handshake communication and process flow knowledge graph reasoning to adapt to the flexible and changing pipeline topology in modular assembly scenarios, dynamically determining the influence coefficients of each neighboring node. Unlike traditional methods that rely on offline fixed models, this invention uses node functional role labels, pipeline access locations, and process material flow relationships to finely classify neighboring nodes into upstream and downstream on the same branch, cross-branch related, or independent branches. It accurately calculates differentiated weights based on topological distance and attenuation factors, ensuring that the influence coefficients truly reflect the strength distribution of pipeline pressure coupling. Simultaneously, by receiving pre-broadcast action intent commands from neighboring nodes before execution, and extracting the expected target frequency and action timestamp, this invention achieves early prediction of neighboring pressure fluctuations, shifting the timing of feedforward compensation from "passive response after disturbance" to "active intervention before disturbance," fundamentally eliminating control lag. Combined with a weighted synthesis method that accumulates pressure fluctuation curves according to positive pressure impact or negative pressure suction classification, the resulting feedforward compensation signal can accurately characterize the amplitude and direction of coupled disturbances, providing a reliable basis for subsequent compensation.

[0015] 2. This invention introduces virtual inertia buffer conditioning before applying the compensation signal. Based on the node's own functional role label, it matches the corresponding virtual inertia characteristics and flattens the abrupt components in the feedforward compensation signal along the time axis. A smooth transition curve with an initial increase followed by a decrease in slope is generated to replace the original abrupt signal, effectively avoiding pressure spikes and mechanical shocks caused by compensation overshoot. This mechanism gives nodes with different roles differentiated response flexibility, ensuring the rapid adjustment needs of the suction source node while preventing system oscillations at the terminal gas consumption point due to overcompensation, achieving a reasonable trade-off between stiffness and flexibility. Simultaneously, it transforms fragmented independent control into collaborative control based on intent sharing and role adaptation, significantly suppressing pipeline pressure fluctuations and chain oscillations, reducing overall system energy consumption, and extending the service life of actuators such as fans and valves. It is particularly suitable for industrial negative pressure applications with frequent start-stop cycles and sudden load changes. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating a collaborative control method for a negative pressure system based on modular assembly, provided in one embodiment of the present invention. Figure 2 A functional block diagram of a negative pressure system collaborative control system based on modular assembly is provided in one embodiment of the present invention; The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0017] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0018] This application provides a modularly assembled negative pressure system collaborative control method. The executing entity of this modularly assembled negative pressure system collaborative control method includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the modularly assembled negative pressure system collaborative control method can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0019] Reference Figure 1 The diagram shown is a flowchart illustrating a modular assembly-based collaborative control method for a negative pressure system according to an embodiment of the present invention. In this embodiment, the modular assembly-based collaborative control method for a negative pressure system includes: S1. Adjacent nodes in the negative pressure pipeline network communicate by handshaking, and determine the influence coefficient of each node based on the process relationship between the nodes after communication. In this embodiment of the invention, adjacent nodes in the negative pressure pipeline network perform handshake communication, and determine the influence coefficient of each node based on the process relationship between the nodes after communication, including: Adjacent nodes exchange their unique identifiers and functional role labels that represent their role in the process flow, wherein the functional role label identifies at least a suction source or a terminal gas consumption point. Based on the neighbor functional role tags obtained from the exchange, query the pre-established process flow ontology knowledge graph, and deduce the upstream and downstream logical relationships between this node and neighboring nodes. Based on the upstream and downstream logical relationship, an influence weight coefficient is calculated for each neighbor node.

[0020] Based on the upstream and downstream logical relationship, an influence weight coefficient is calculated for each neighbor node, including: Based on the upstream and downstream logical relationship, neighboring nodes are classified into upstream and downstream nodes on the same branch, cross-branch related nodes, or independent branch nodes, and the influence weights of the corresponding levels are assigned to neighboring nodes of different categories.

[0021] The adjacent nodes in the negative pressure pipeline network communicate by handshaking, and the influence coefficient of each node is determined based on the process relationship between the nodes after communication. This also includes: During the handshake communication phase, the network access location description information reported by neighboring nodes is received; By combining the functional role label of this node with the pipeline access location description information of the neighboring node, it is determined whether there is a direct process material flow relationship between the two nodes. Based on the determination that there is a direct process material flow relationship, the neighboring nodes from which process materials flow out are marked as upstream influence sources, and the neighboring nodes from which process materials flow in are marked as downstream influence objects. The upstream influence sources are assigned an influence coefficient that is higher than that of the downstream influence objects.

[0022] The formula for calculating the influence coefficient is as follows: ; In the formula, For nodes Neighboring nodes The influence coefficient of the allocation Neighboring nodes When nodes are classified as upstream or downstream nodes of the same branch, the baseline value is used. For nodes with neighboring nodes Topological distance, For nodes with neighboring nodes Topological distance, For nodes The set of all neighboring nodes, is the ordinal number of the neighboring node. It is a natural constant. Neighboring nodes When nodes are classified as upstream or downstream nodes of the same branch, the baseline value is used. This is a preset attenuation factor based on the physical characteristics of the pipeline network.

[0023] In the negative pressure pipeline network, adjacent nodes establish handshake communication, and the communication link remains stable. Each node sends its unique and non-repeatable identifier to its neighboring nodes, and simultaneously transmits its functional role label, which accurately represents its specific role in the overall process of the negative pressure pipeline network. Neighboring nodes send their unique identifiers and functional role labels to this node using the same transmission method. This node receives and verifies the above information. The functional role label strictly defines the node identifier as a suction source or terminal gas consumption point, ensuring that nodes can accurately identify each other's position and responsibilities in the process, providing complete and accurate basic information for subsequent relationship reasoning.

[0024] This node uses the functional role tags of adjacent nodes as the core query basis. It performs targeted matching and retrieval operations in the pre-constructed, rule-complete, and clearly defined process flow ontology knowledge graph. According to the pre-set node role correspondence rules within the knowledge graph, it compares the graph entity entries corresponding to the functional role tags of this node and adjacent nodes in turn. Based on the pre-set logical pointing and hierarchical constraints between entities in the graph, it directly derives the clear and unique upstream and downstream logical relationship between this node and adjacent nodes in the negative pressure pipeline process. The derivation process is executed entirely according to the inherent rules of the knowledge graph, and stable and reliable relationship results can be obtained without external assistance.

[0025] Based on the established upstream and downstream logical relationships between this node and its neighbors, this node determines the degree of influence of each neighboring node according to a pre-set and uniformly executed node influence determination standard. Based on the direction, scope, and level of influence determined by the upstream and downstream logical relationships, the influence weight coefficient corresponding to each neighboring node is directly determined. All influence weight coefficients of neighboring nodes are generated according to the same standard, and each coefficient maintains a high degree of matching with the logical relationship between the corresponding node, ensuring that the influence weight coefficient has uniqueness and accuracy.

[0026] Based on the established upstream and downstream logical relationships between this node and its neighboring nodes, and according to the pre-set branch attribution rules, the pipeline connection path and process attribution range of each neighboring node in the negative pressure pipeline network are checked one by one. Each neighboring node is accurately classified into the corresponding category. Neighboring nodes that are on the same pipeline process path and have direct upstream and downstream connections are classified as upstream and downstream nodes of the same branch. Neighboring nodes that are indirectly related to the process through other branches and have no direct upstream and downstream connections are classified as cross-branch related nodes. Neighboring nodes that have no process path connection with this node and operate independently are classified as independent branch nodes. All neighboring nodes are uniquely and clearly classified, without omissions or overlaps.

[0027] For the three types of nodes that have been classified—upstream and downstream nodes on the same branch, related nodes across branches, and independent branch nodes—the system assigns corresponding level influence weights to neighboring nodes of different categories according to a pre-set and uniformly executed hierarchical weight allocation rule. The highest level influence weight is assigned to upstream and downstream nodes on the same branch, the medium level influence weight is assigned to related nodes across branches, and the lowest level influence weight is assigned to independent branch nodes. The weight allocation process is strictly executed one by one according to the correspondence between the classification results and the hierarchy. Each neighboring node obtains an influence weight that is completely matched with its own classification. The weight allocation result is stable and unique, and maintains a high degree of consistency with the node classification and the upstream and downstream logical relationship.

[0028] During the entire handshake communication process, adjacent nodes in the negative pressure pipeline network maintain a stable communication link. During the communication phase, this node continuously receives pipeline access location description information reported in real time by neighboring nodes through the communication link. This information completely records the specific access pipeline number, access segment location, access port number, and relative position with the surrounding pipeline structure of the neighboring node in the negative pressure pipeline network. This node verifies and stores the received pipeline access location description information field by field to ensure that the information is complete and error-free, providing a true and accurate location basis for subsequent node relationship judgment. All access location information comes from the active reporting of neighboring nodes and is not externally collected or modified.

[0029] The node's pre-determined functional role label is comprehensively compared with the received and stored pipeline access location description information of neighboring nodes. According to the pre-set process material flow path determination rules of the negative pressure pipeline network, the position association and functional matching of the two nodes are verified item by item. The functional role label clarifies the node's suction or gas consumption attribute in the process flow, and the pipeline access location description information confirms the physical connection status of the two nodes in the pipeline network structure. Based on the positional connectivity and functional matching, a comprehensive judgment is made on whether there is a direct process material flow relationship between the two nodes. The judgment process strictly follows the pre-set rules to obtain a unique and definite judgment result.

[0030] For two nodes that are determined to have a direct process material flow relationship, a marking operation is performed according to the actual transmission direction of the process material. Neighboring nodes from which the process material flows out are uniformly marked as upstream influencing sources, and neighboring nodes from which the process material flows in are uniformly marked as downstream influencing objects. After marking is completed, the upstream influencing source is assigned an influence coefficient higher than that of the downstream influencing object according to the preset influence coefficient level allocation rules. The process of assigning the influence coefficient is strictly executed according to the marking type. The influence coefficient corresponding to the upstream influencing source is always kept higher than that of the downstream influencing object. The assignment result is completely matched with the node marking and the material flow direction, without deviation or reversal.

[0031] node Neighboring nodes In the process of calculating the influence coefficient of allocation, neighboring nodes The corresponding parameters are directly determined by the previous node relationship classification results, when neighboring nodes When nodes are determined to be upstream or downstream nodes on the same branch, this parameter takes a preset baseline value. This parameter is used to characterize neighboring nodes. The role and weight in the process flow provide a classification basis for the allocation of influence coefficients.

[0032] node with neighboring nodes The topological distance is determined by traversing the nodes in the negative pressure pipeline network. to neighboring nodes The number of nodes or pipe segments contained in the connected pipeline path is counted. This parameter is used to quantify the degree of connection between two nodes in the physical structure of the pipeline network, and serves as a basis for the attenuation of influence with distance.

[0033] node with neighboring nodes The topological distance is also calculated by traversing the nodes. to each neighbor node The connected path statistics are obtained and used to calculate the weighted influence value of other neighboring nodes.

[0034] node The set of all neighboring nodes is composed of nodes The information obtained from all adjacent nodes during the handshake communication phase is summarized and used to determine the coverage of the denominator term in the influence coefficient calculation.

[0035] The attenuation factor, preset based on the physical characteristics of the pipeline network, is determined by a combination of the negative pressure transmission characteristics, pipeline resistance characteristics, and node distribution characteristics of the pipeline network. It is used to control the rate at which the degree of influence decreases as the topological distance increases.

[0036] The natural constant is a fixed mathematical constant used to construct the exponential trend of the influence decaying with distance.

[0037] Neighbor nodes The corresponding parameters are from the neighboring nodes. The classification results are directly determined, and the upstream and downstream nodes on the same branch take the benchmark value to represent the role weight of each neighbor node and participate in the calculation of the weighted influence value of all neighbor nodes.

[0038] node Select each neighbor node sequentially from the set of neighbor nodes. Obtain the classification result of the neighboring node and determine the corresponding parameter values, while also obtaining the node... with neighboring nodes The topological distance is used to calculate the weighted influence value of the neighboring node by combining the parameter values, topological distance, and attenuation factor. Then, the node... Iterate through all neighbor nodes in the set of neighbor nodes. Repeat the above calculation process to obtain each neighbor node. The weighted influence value is calculated, and the weighted influence values ​​of all neighboring nodes are summed to obtain the sum of the denominator terms. Finally, the neighboring nodes are... Dividing the weighted influence value of a node by the sum of the weighted influence values ​​of all its neighboring nodes yields the node's weighted influence value. Neighboring nodes The influence coefficient is assigned, and this calculation process is performed independently for each neighbor node to ensure that each neighbor node receives a unique corresponding influence coefficient.

[0039] This formula is used to implement nodes. The quantitative allocation of the influence of each neighboring node is achieved by combining the classification attributes of neighboring nodes with the topological distance between nodes, and determining the influence coefficient of each neighboring node in the form of a weighted ratio. This allows for the precise differentiation of the influence of different neighboring nodes on the node. The degree of influence is determined by the fact that upstream and downstream nodes on the same branch are given higher influence weights because they are directly involved in material transmission. The closer the nodes are in the topology, the higher their influence. At the same time, the attenuation factor is used to control the attenuation rate of influence with distance, ensuring that the distribution of influence coefficients is both in line with the process role of the nodes and in line with the physical structure characteristics of the pipeline network. This provides an accurate quantitative basis for the prediction of pressure disturbances between subsequent nodes and for feedforward compensation.

[0040] The beneficial effects include stable information interaction through handshake communication between adjacent nodes in the negative pressure pipeline network, accurate inference of upstream and downstream logical relationships based on functional role tags and process flow ontology knowledge graph, precise classification of neighboring nodes and collection of access location information, and determination of standardized and accurate influence weight coefficients for each adjacent node in a unified quantitative manner by combining node classification attributes and topological distance. This clearly defines the process positioning, direct process material flow relationship, and mutual influence degree of each node, solving the problem that traditional static weights cannot adapt to differences in node roles and distance attenuation characteristics. This makes the determination of node influence relationships more accurate and reasonable, improves the pertinence and effectiveness of node process control, provides reliable quantitative basis for subsequent pressure fluctuation prediction and feedforward compensation, strengthens the collaborative cooperation and overall operational coordination among pipeline network nodes, ensures the continuous and stable operation of the negative pressure delivery process, improves the controllability, reliability, and execution efficiency of the overall pipeline network operation, and provides comprehensive and accurate node relationship support for pipeline network process scheduling, fault location, and operation optimization.

[0041] S2. Based on the action intentions of all neighboring nodes around this node, predict the pressure fluctuation curves of the neighboring nodes, and weight and superimpose the influence coefficients according to the pressure fluctuation curves to obtain the feedforward compensation signal of this node. In this embodiment of the invention, predicting the pressure fluctuation curves of neighboring nodes based on the action intentions of all neighboring nodes surrounding the current node includes: This node receives the action intent instructions pre-broadcast by each neighboring node before executing the control action, and extracts the expected value of the target frequency to be adjusted by the neighboring node and the action execution timestamp from the action intent instructions; Based on the expected target frequency, determine the wind turbine speed change range of neighboring nodes, and align the actions of neighboring nodes in time sequence by combining the action execution timestamps. For each neighboring node, based on the fan speed change range, the corresponding pipeline pressure response characteristic curve of the neighboring node is called to generate the pressure fluctuation curve that the neighboring node is expected to cause at the outlet of this node.

[0042] The step of weighting and superimposing the influence coefficients based on the pressure fluctuation curve to obtain the feedforward compensation signal for this node includes: This node obtains the pressure fluctuation curves and influence coefficients of each neighboring node to establish a mapping table between pressure fluctuation curves and influence coefficients. The type of disturbance caused by the actions of neighboring nodes to the negative pressure of this node is determined based on the direction of the pressure fluctuation curve. The disturbance type is classified as positive pressure impact or negative pressure suction. After expanding the pressure fluctuation curve along the time axis, at each time sampling point, the pressure fluctuation value of each neighboring node at the current time is multiplied by its corresponding influence coefficient to obtain the weighted fluctuation component. Then, the weighted fluctuation components of all neighboring nodes are accumulated according to the disturbance type to obtain the feedforward compensation signal.

[0043] This node maintains a stable information exchange state with all neighboring nodes. Before each neighboring node executes a control action, it fully receives the action intent instructions actively broadcast by each neighboring node. It then performs field-by-field parsing and content extraction on the received action intent instructions, accurately separating the target frequency expectation value corresponding to the neighboring node's planned adjustment from all the information contained in the instruction. At the same time, it extracts the action execution timestamp corresponding to the neighboring node's planned control action. The extraction process is completed one by one according to the preset information splitting rules, ensuring that the target frequency expectation value and action execution timestamp information are complete and accurate and match the corresponding neighboring nodes one by one, without any information confusion or loss.

[0044] Based on the extracted target frequency expected value corresponding to the neighbor node, and according to the preset frequency and speed correspondence judgment rule, the increase or decrease range of the fan speed during the execution of the control action by the neighbor node is directly determined. After the determination is completed, the fan speed change range is bound and associated with the action execution timestamp corresponding to the same neighbor node. According to the order marked by the action execution timestamp, the action execution times of all neighbor nodes are uniformly arranged so that the actions of different neighbor nodes are in corresponding standard positions on the time axis, and the timing alignment of all neighbor node actions is completed. The timing alignment result is completely consistent with the action execution timestamp.

[0045] For each neighboring node that has completed time alignment, the fan speed change amplitude corresponding to that node is used as the reference. The pipeline pressure response characteristic curve corresponding to that neighboring node is retrieved from the pre-stored node-specific data set. According to the correspondence between the fan speed change amplitude and the pipeline pressure response characteristic curve, the pressure change characteristics and time change characteristics are matched segment by segment. The curve characteristics are combined with the speed change amplitude for complete deduction. The pressure fluctuation curve that the neighboring node is expected to cause at the outlet position of this node after performing the control action is generated. Each neighboring node generates a unique corresponding pressure fluctuation curve, and the content of the curve is completely matched with the fan speed change amplitude.

[0046] This node obtains the pressure fluctuation curves corresponding to each neighboring node from the local storage unit and previous interaction records. At the same time, it obtains the influence coefficients of each neighboring node that have been determined and assigned values ​​in the previous stage. The obtained pressure fluctuation curves and influence coefficients are matched one by one according to the unique identifier of the neighboring node. According to the preset table construction rules, the corresponding pressure fluctuation curves and influence coefficients are filled into the corresponding positions in the table in sequence to form a complete and unique mapping table of correspondence between pressure fluctuation curves and influence coefficients. Each entry in the mapping table clearly points to a single neighboring node, and there are no duplicate or missing correspondences.

[0047] This node retrieves the pressure fluctuation curves of each neighboring node in the corresponding relationship mapping table, examines and judges the numerical change trend of each pressure fluctuation curve, determines the fluctuation direction based on the overall upward or downward trend of the pressure fluctuation curve, and classifies the fluctuation direction according to the preset disturbance type classification standard. The fluctuation direction with an upward trend in pressure value is judged as a positive pressure impact type disturbance, and the fluctuation direction with a downward trend in pressure value is judged as a negative pressure suction type disturbance. Each pressure fluctuation curve corresponds to a uniquely determined disturbance type, and the disturbance type classification result is completely consistent with the fluctuation direction of the pressure fluctuation curve.

[0048] This node fully expands the pressure fluctuation curves corresponding to all neighboring nodes along a unified time axis, so that each curve presents the pressure change state at the same time scale. At each time sampling point on the time axis, the pressure fluctuation value corresponding to each neighboring node at the current time sampling point is extracted one by one. The pressure fluctuation value is then combined with the influence coefficient matched by the corresponding relationship mapping table of the same neighboring node to obtain the weighted fluctuation component of the current neighboring node at the current time sampling point. The weighted fluctuation components generated by all neighboring nodes are classified into two disturbance types: positive pressure impact and negative pressure suction. All weighted fluctuation components under the same type are accumulated sequentially to finally form the feedforward compensation signal required by this node.

[0049] The beneficial effects are as follows: by receiving the action intention instructions of neighboring nodes and extracting key information, the amplitude of fan speed change and timing alignment results are determined. A pressure fluctuation curve is generated by combining the pipeline pressure response characteristic curve. Then, a mapping table of the correspondence between the pressure fluctuation curve and the influence coefficient is constructed. This accurately distinguishes the type of disturbance and completes weighted superposition processing. It can predict the pressure change trend at the outlet of this node in advance before neighboring nodes execute control actions, accurately generate feedforward compensation signals that fit the actual operating state of the pipeline network, effectively offsetting the pressure disturbances caused by neighboring node actions. This improves the foresight, response speed, and adjustment accuracy of negative pressure pipeline network pressure control, avoids the impact of sudden pressure changes and large fluctuations on pipeline network operation, ensures the continuous stability of the negative pressure state of this node and the overall smooth and reliable operation of the pipeline network, and provides accurate prediction and compensation basis for pipeline network collaborative control. This enhances the coordination of control cooperation between nodes, improves the stability, safety, intelligence level, and reliability of the entire negative pressure transmission system, and strengthens the overall operating efficiency and safety stability of the pipeline network.

[0050] S3. Before the feedforward compensation signal is superimposed onto the local node, virtual inertia is calculated. In this embodiment of the invention, the step of performing virtual inertia calculation before the feedforward compensation signal is superimposed onto the current node includes: This node obtains its own functional role label determined in the handshake communication, and determines the target virtual inertia characteristic to be adopted by this node based on the correspondence between the functional role label and the preset virtual inertia characteristic. The feedforward compensation signal is input to a signal buffer with the target virtual inertia characteristics, and the signal buffer performs time-level flattening on the abrupt amplitude of the feedforward compensation signal. The signal buffer outputs a flattened feedforward compensation signal, which is used as the compensation signal to be superimposed on this node.

[0051] This node retrieves its own unique functional role label from the information set confirmed and stored during the previous handshake communication process. This functional role label clearly identifies the specific responsibility of this node as a suction source or terminal gas consumption point in the negative pressure pipeline process. The retrieved functional role label is matched one by one with the pre-set correspondence between functional role labels and virtual inertia characteristics. According to the successfully matched correspondence, the target virtual inertia characteristic that this node should use in the virtual inertia processing stage is directly locked. The matching process is strictly executed according to the preset correspondence rules to obtain a unique target virtual inertia characteristic that is fully compatible with the function of this node.

[0052] This node transmits the feedforward compensation signal obtained through weighted superposition processing to a signal buffer stage with predetermined processing capabilities. Before receiving the signal, the signal buffer stage has completed the parameter and operation rules configuration according to the determined target virtual inertia characteristics, and has the signal adjustment capability that is completely consistent with the target virtual inertia characteristics. After the feedforward compensation signal enters the signal buffer stage, the stage identifies the abrupt amplitude of the feedforward compensation signal segment by segment according to the signal change constraint rules corresponding to the target virtual inertia characteristics, and then uses a gradual transition method in the time dimension to uniformly flatten the abrupt amplitude, eliminating the rapidly changing parts of the signal.

[0053] After completing the amplitude flattening process of the feedforward compensation signal, the signal buffering stage maintains the integrity and continuity of the signal during the processing and outputs the fully flattened feedforward compensation signal stably. The output signal no longer contains abrupt amplitude changes and the trend of change is gentle. This signal is directly used as the compensation signal waiting to be superimposed on this node. The characteristics of the compensation signal waiting to be superimposed on this node are completely consistent with the target virtual inertia characteristics and can be directly used for subsequent compensation superposition operations of this node, ensuring that there are no abnormalities or losses in the signal transmission and use process.

[0054] The beneficial effects are that by matching the functional role labels to determine the target virtual inertia characteristics and by using the signal buffering stage to flatten the feedforward compensation signal, it is possible to effectively avoid the impact of sudden changes in the feedforward compensation signal on the control system of this node, making the compensation signal change more gradual and stable, improving the smoothness and reliability of the negative pressure regulation process of this node, preventing drastic fluctuations in pipeline pressure, and ensuring that the compensation signal is highly compatible with the functional role of this node, enhancing the pertinence and effectiveness of the compensation operation, ensuring the overall stable operation of the negative pressure pipeline network, extending the stable operation cycle of the node control equipment, and improving the anti-interference capability and operational safety of the pipeline network system.

[0055] S4. Compensate the current node according to the corrected cooperative control quantity.

[0056] In this embodiment of the invention, after the signal buffering stage outputs a flattened feedforward compensation signal as the signal to be superimposed onto the compensation signal of this node, the method further includes: This node monitors the rate of change of the feedforward compensation signal in real time. When the rate of change exceeds the preset stable change threshold, it determines that there is a sudden change component in the feedforward compensation signal. Extract the magnitude of the mutation component and redistribute the magnitude of the mutation according to the time axis to generate a smooth compensation transition curve with the change slope first increasing and then decreasing. The abrupt component in the original feedforward compensation signal is replaced by the smooth compensation transition curve to form a gradually changing feedforward compensation signal, allowing the node to have a buffer time before performing the compensation action.

[0057] This node outputs the flattened feedforward compensation signal in the signal buffer stage and determines the compensation signal to be superimposed on this node. It continuously collects and tracks the numerical changes of the feedforward compensation signal, calculates the change amplitude of the feedforward compensation signal between adjacent time moments at fixed time intervals, and obtains the real-time change rate of the feedforward compensation signal. The obtained real-time change rate is directly compared with the pre-set and stored stable change threshold. When the value of the real-time change rate is greater than the value range limited by the stable change threshold, it is directly determined that there is an abrupt component that has not been completely eliminated in the current feedforward compensation signal. The determination result is unique and clear.

[0058] After determining that there is abrupt change in the feedforward compensation signal, this node accurately locates the interval with the abrupt change in the feedforward compensation signal. It then extracts the overall change amplitude corresponding to the abrupt change component from the located abrupt change interval. The extraction process fully preserves the peak value and duration information of the abrupt change component. The extracted change amplitude is then redistributed uniformly and orderly according to a unified time axis. During the redistribution process, the curve profile is constructed by first slowly increasing the change rate and then gradually decreasing the change rate. This makes the generated curve have the characteristic of the change slope first increasing and then decreasing, and finally forms a complete and continuous smooth compensation transition curve. The curve shape and change amplitude are perfectly matched.

[0059] This node compares the generated smooth compensation transition curve with the original feedforward compensation signal segment by segment to locate the specific location and range of abrupt changes in the original feedforward compensation signal. It then replaces the corresponding abrupt change portion of the original feedforward compensation signal with the complete smooth compensation transition curve. The replacement process maintains the continuous connection between the preceding and following intervals of the signal, without any breaks or jumps. After the replacement is completed, a feedforward compensation signal with a uniform and gentle overall trend and no abrupt changes is formed. This signal can reserve sufficient buffer time for this node to perform compensation actions, avoiding the impact caused by the node device's rapid response due to signal abrupt changes.

[0060] The beneficial effects are that by monitoring the rate of change of the feedforward compensation signal in real time and identifying abrupt changes, the amplitude of the changes is extracted to generate a smooth compensation transition curve for replacement, which can further eliminate the influence of abrupt changes in the compensation signal, making the changes in the compensation signal more gradual and orderly, significantly reducing the instantaneous impact of abrupt changes on the control mechanism of this node, ensuring that the node performs compensation actions smoothly and stably, while providing sufficient buffer time for node control, improving the overall operational stability of the negative pressure pipeline network and the service life of equipment, and enhancing the system's anti-disturbance capability and control reliability.

[0061] like Figure 2 The diagram shown is a functional block diagram of a negative pressure system collaborative control system based on modular assembly, provided by an embodiment of the present invention.

[0062] The modularly assembled negative pressure system collaborative control system 100 described in this invention can be installed in an electronic device. Depending on the functions implemented, the modularly assembled negative pressure system collaborative control system 100 may include a topology handshake and weight calculation module 101, an intent prediction and feedforward synthesis module 102, a virtual inertia buffer conditioning module 103, and a collaborative compensation execution module 104. The module described in this invention can also be called a unit, referring to a series of computer program segments that can be executed by the processor of an electronic device and perform a fixed function, stored in the memory of the electronic device.

[0063] In this embodiment, the functions of each module / unit are as follows: The topology handshake and weight calculation module 101 is used for adjacent nodes in the negative pressure pipeline to perform handshake communication, and to determine the influence coefficient of each node based on the process relationship between the nodes after communication. The intent prediction and feedforward synthesis module 102 is used to predict the pressure fluctuation curve of the neighboring nodes based on the action intent of all neighboring nodes around the current node, and to weight and superimpose the influence coefficients based on the pressure fluctuation curves to obtain the feedforward compensation signal of the current node. The virtual inertia buffer conditioning module 103 is used to perform virtual inertia before the feedforward compensation signal is superimposed on the local node. The collaborative compensation execution module 104 is used to compensate the local node according to the corrected collaborative control quantity.

[0064] In the several embodiments provided by this invention, it should be understood that the disclosed methods and systems can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0065] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0066] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0067] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0068] This application embodiment can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0069] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A collaborative control method for a negative pressure system based on modular assembly, characterized in that, The method includes: S1. Adjacent nodes in the negative pressure pipeline network communicate by handshaking, and determine the influence coefficient of each node based on the process relationship between the nodes after communication. S2. Based on the action intentions of all neighboring nodes around this node, predict the pressure fluctuation curves of the neighboring nodes, and weight and superimpose the influence coefficients according to the pressure fluctuation curves to obtain the feedforward compensation signal of this node. S3. Before the feedforward compensation signal is superimposed onto the local node, virtual inertia is calculated. S4. Compensate the current node according to the corrected cooperative control quantity.

2. The method for coordinated control of a negative pressure system based on modular assembly as described in claim 1, characterized in that, Adjacent nodes in the negative pressure pipeline network communicate via handshake and determine the influence coefficient of each node based on the flow relationship between nodes after communication, including: Adjacent nodes exchange their unique identifiers and functional role labels that represent their role in the process flow, wherein the functional role label identifies at least a suction source or a terminal gas consumption point. Based on the neighbor functional role tags obtained from the exchange, query the pre-established process flow ontology knowledge graph, and deduce the upstream and downstream logical relationships between this node and neighboring nodes. Based on the upstream and downstream logical relationship, an influence weight coefficient is calculated for each neighbor node.

3. The method for coordinated control of a negative pressure system based on modular assembly as described in claim 2, characterized in that, Based on the upstream and downstream logical relationship, an influence weight coefficient is calculated for each neighbor node, including: Based on the upstream and downstream logical relationship, neighboring nodes are classified into upstream and downstream nodes on the same branch, cross-branch related nodes, or independent branch nodes, and the influence weights of the corresponding levels are assigned to neighboring nodes of different categories.

4. The method for coordinated control of a negative pressure system based on modular assembly as described in claim 3, characterized in that, The adjacent nodes in the negative pressure pipeline network communicate by handshaking, and the influence coefficient of each node is determined based on the process relationship between the nodes after communication. This also includes: During the handshake communication phase, the network access location description information reported by neighboring nodes is received; By combining the functional role label of this node with the pipeline access location description information of the neighboring node, it is determined whether there is a direct process material flow relationship between the two nodes. Based on the determination that there is a direct process material flow relationship, the neighboring nodes from which process materials flow out are marked as upstream influence sources, and the neighboring nodes from which process materials flow in are marked as downstream influence objects. The upstream influence sources are assigned an influence coefficient that is higher than that of the downstream influence objects.

5. The method for coordinated control of a negative pressure system based on modular assembly as described in claim 4, characterized in that, The formula for calculating the influence coefficient is as follows: ； In the formula, For nodes Neighboring nodes The influence coefficient of the allocation Neighboring nodes When nodes are classified as upstream or downstream nodes of the same branch, the baseline value is used. For nodes with neighboring nodes Topological distance, For nodes with neighboring nodes Topological distance, For nodes The set of all neighboring nodes, is the ordinal number of the neighboring node. It is a natural constant. Neighboring nodes When nodes are classified as upstream or downstream nodes of the same branch, the baseline value is used. This is a preset attenuation factor based on the physical characteristics of the pipeline network.

6. The method for coordinated control of a negative pressure system based on modular assembly as described in claim 1, characterized in that, The step of predicting the pressure fluctuation curves of neighboring nodes based on the action intentions of all neighboring nodes around this node includes: This node receives the action intent instructions pre-broadcast by each neighboring node before executing the control action, and extracts the expected value of the target frequency to be adjusted by the neighboring node and the action execution timestamp from the action intent instructions; Based on the expected target frequency, determine the wind turbine speed change range of neighboring nodes, and align the actions of neighboring nodes in time sequence by combining the action execution timestamps. For each neighboring node, based on the fan speed change, the corresponding pipeline pressure response characteristic curve of the neighboring node is called to generate the pressure fluctuation curve that the neighboring node is expected to cause at the outlet of this node.

7. The method for coordinated control of a negative pressure system based on modular assembly as described in claim 6, characterized in that, The step of weighting and superimposing the influence coefficients based on the pressure fluctuation curve to obtain the feedforward compensation signal for this node includes: This node obtains the pressure fluctuation curves and influence coefficients of each neighboring node to establish a mapping table between pressure fluctuation curves and influence coefficients. The type of disturbance caused by the actions of neighboring nodes to the negative pressure of this node is determined based on the direction of the pressure fluctuation curve. The disturbance type is classified as positive pressure impact or negative pressure suction. After expanding the pressure fluctuation curve along the time axis, at each time sampling point, the pressure fluctuation value of each neighboring node at the current time is multiplied by its corresponding influence coefficient to obtain the weighted fluctuation component. Then, the weighted fluctuation components of all neighboring nodes are accumulated according to the disturbance type to obtain the feedforward compensation signal.

8. The method for coordinated control of a negative pressure system based on modular assembly as described in claim 1, characterized in that, The step of performing virtual inertia before the feedforward compensation signal is superimposed onto the current node includes: This node obtains its own functional role label determined in the handshake communication, and determines the target virtual inertia characteristic to be adopted by this node based on the correspondence between the functional role label and the preset virtual inertia characteristic. The feedforward compensation signal is input to a signal buffer with the target virtual inertia characteristics, and the signal buffer performs time-wise flattening processing on the abrupt amplitude of the feedforward compensation signal. The signal buffer outputs a flattened feedforward compensation signal, which is used as the compensation signal to be superimposed on this node.

9. The method for coordinated control of a negative pressure system based on modular assembly as described in claim 8, characterized in that, The signal buffering stage outputs a flattened feedforward compensation signal, which is then superimposed onto the compensation signal of this node. The system also includes: This node monitors the rate of change of the feedforward compensation signal in real time. When the rate of change exceeds the preset stable change threshold, it determines that there is a sudden change component in the feedforward compensation signal. Extract the magnitude of the mutation component and redistribute the magnitude of the mutation according to the time axis to generate a smooth compensation transition curve with the change slope first increasing and then decreasing. The abrupt component in the original feedforward compensation signal is replaced by the smooth compensation transition curve to form a gradually changing feedforward compensation signal, allowing the node to have a buffer time before performing the compensation action.

10. A modular assembly-based negative pressure system collaborative control system, characterized in that, For implementing the modular assembly-based negative pressure system collaborative control method of claim 1, the system includes: The topology handshake and weight calculation module is used for adjacent nodes in a negative pressure pipeline network to perform handshake communication and determine the influence coefficient of each node based on the process relationship between nodes after communication. The intention prediction and feedforward synthesis module is used to predict the pressure fluctuation curves of neighboring nodes based on the action intentions of all neighboring nodes around the current node, and to weight and superimpose the influence coefficients based on the pressure fluctuation curves to obtain the feedforward compensation signal of the current node. The virtual inertia buffer conditioning module is used to perform virtual inertia before the feedforward compensation signal is superimposed onto the local node. The collaborative compensation execution module is used to compensate the local node based on the corrected collaborative control quantity.

Images