An open search-regulation mechanism based on the ΔN–ΔD model that reduces combinatorial explosion.
Metric: CR_nodes = baseline_nodes / dn_nodes
Results:
- Mean CR = 8.394×
- Median CR = 8.467×
- Min CR = 5.431×
- Max CR = 11.308×
This means CCE, driven by the ΔN–ΔD control model, reduces the number of explored nodes by approximately 8× on average compared to the baseline.
ΔN–ΔD model → CCE regulator → Backends ΔN–ΔD is a formal dynamic framework describing structural evolution through two orthogonal parameters: nonequilibrium (ΔN) and duality (ΔD).
core/— CCE core implementationbackends/— stress-test and decoding demovisualizer/— pulse plots (ΔN, ΔD over time)experiments/— generated CSV and PNG resultsdocs/— model formalization
The structural dynamics implemented in this repository are based on the following equation:
dS/dt = α(ΔN, ΔD) · ΔN + β · ΔD
where:
- ΔN — nonequilibrium (external gradient),
- ΔD — structural duality (internal divergence),
- α(ΔN, ΔD) — nonlinear system response,
- β — structural contribution coefficient.
Install dependencies:
pip install -r requirements.txt
Run the batch test (30 runs):
python run_search_stress.py
Results are saved in the experiments/ folder:
-
search_stress_results.csv
-
CR over seeds plot
-
baseline vs DN scatter plot
-
pulse plots
The ΔN–ΔD theoretical framework and the CCE architecture presented in this repository are original works by Zhengis Tileubay and are covered by an international patent application:
- Patent Application: International Patent Publication WO/2026/155638 (Method for Controlling the Dynamics of a Complex System, PCT/KZ2026/000001). © 2026 Zhengis Tileubay. Source code and documentation are licensed for research and non-commercial use only. Commercial licensing or technology transfer agreements are available upon request. For licensing, partnership, or collaboration inquiries, please contact the author.